representation of x rays

• Random article
• Teaching guide
• Privacy & cookies

by Chris Woodford . Last updated: October 31, 2022.

Photo: Once X rays had to be treated like old-fashioned photographs. Now, they're as easy to study and store as digital photographs on computer screens. Photo by Kasey Zickmund courtesy of U.S. Air Force.

What are X rays?

Artwork: The electromagnetic spectrum, with the X-ray band highlighted in yellow over toward the right. You can see that X rays have shorter wavelengths, higher frequencies, and higher energy than most other types of electromagnetic radiation, and don't penetrate Earth's atmosphere. Their wavelengths are around the same scale as atomic sizes. Artwork courtesy of NASA (please follow this link for a bigger and clearer version of this image).

Artwork: Lead is a heavy element that you'll find toward the bottom of the periodic table: its atoms contain lots of protons and neutrons, so they're very dense and heavy. Lead is very good at stopping X rays.

What are X rays used for?

Photo: Taking a dental X ray with modern, digital technology. This equipment uses low-power (and therefore safer) X rays and instead of the dentist having to develop an old-fashioned photo, the results show up almost instantly on their computer screen. Photo by Matthew Lotz courtesy of US Air Force .

Photo: A typical CT scanner. The patient lies on the bed, which slides through the hole in the donut-shaped scanner behind. The scanner unit contains one or more rotating X-ray sources and detectors. Photo by Francisco V. Govea II courtesy of US Air Force and Wikimedia Commons .

Photo: Using digital X ray equipment (left) to check the contents of a suspicious package (on the floor, right). Photo by Jonathan Pomeroy courtesy of US Air Force .

Photo: Nondestructive X ray testing is one way to inspect planes without taking them apart. Here, a plane has just been tested in a lead-lined hangar at Randolph US Air Force Base, Texas. The warning signs you can see on the door indicate the potential dangers from the X rays. Photo by Steve Thurow courtesy of US Air Force.

Photo: Studying semiconductor materials with X-ray spectroscopy. Photo by Jim Yost courtesy of US DOE/NREL .

Photo: X-ray image of the Sun produced by the Soft X-ray Telescope (SXT). Photo courtesy of NASA Goddard Space Flight Center (NASA-GSFC) .

How are X rays produced?

How were x rays discovered.

Photo: Wilhelm Röntgen's X-ray photograph of his wife's hand. Note the rings! Photo believed to be in the public domain, courtesy of the National Library of Medicine's Images from the History of Medicine (NLM) collection and the National Institutes of Health .

19th century

20th century.

Illustration: A typical Coolidge tube. Artwork courtesy of the Wellcome Collection published under a Creative Commons (CC BY 4.0) licence .

Photo: The Chandra X-ray telescope just before it was released from the Space Shuttle Columbia on on July 23, 1999. Photo courtesy of NASA/JSC

21st century

If you liked this article..., don't want to read our articles try listening instead, find out more, on this website.

• Electromagnetic spectrum
• Space telescopes

On other sites

Text copyright © Chris Woodford 2009, 2022. All rights reserved. Full copyright notice and terms of use .

Rate this page

Tell your friends, cite this page, more to explore on our website....

• Get the book
• Send feedback

introduction

Electromagnetic waves

Reverse photoelectric effect

X-rays were discovered in 1895 by the German physicist Wilhelm Röntgen (also spelled Roentgen). He received the first Nobel Prize in physics in 1901 "in recognition of the extraordinary services he has rendered by the discovery of the remarkable rays subsequently named after him." Wurzberg Physical-Medical Society, Chairman Albert von Kolliker, whose hand was used to to produce this image, proposed that this new form of radiation be called "Röntgen's Rays". Röntgen had a different idea.

It is seen, therefore, that some agent is capable of penetrating black cardboard which is quite opaque to ultra-violet light, sunlight, or arc-light. It is therefore of interest to investigate how far other bodies can be penetrated by the same agent. It is readily shown that all bodies possess this same transparency, but in very varying degrees. For example, paper is very transparent; the fluorescent screen will light up when placed behind a book of a thousand pages; printer's ink offers no marked resistance…. A piece of sheet aluminium, 15 mm. thick, still allowed the X-rays (as I will call the rays , for the sake of brevity) to pass, but greatly reduced the fluorescence. Glass plates of similar thickness behave similarly; lead glass is, however, much more opaque than glass free from lead…. If the hand be held before the fluorescent screen, the shadow shows the bones darkly, with only faint outlines of the surrounding tissues. Wilhelm Röntgen, 1895

Röntgen appears to have always capitalized the x. I prefer to use lowercase, since the rays are purposely not named after any person, any place, or anything.

Warning: don't try this at home. Don't try this anywhere!

The retina of the eye is quite insensitive to these rays: the eye placed close to the apparatus sees nothing. It is clear from the experiments that this is not due to want of permeability on the part of the structures of the eye. Wilhelm Röntgen, 1895

1912: Walter Friedrich and Paul Knipping diffract x-rays in zinc blende

1912: Max von Laue suggests using lattice solids to diffract x-rays

1913: William Bragg and Lawrence Bragg work out the Bragg condition for strong x-ray reflection

1922: Arthur Compton studies x-ray photon scattering by electrons

Roentgen/Gas-Filled Tubes

The earliest x-ray tubes were filled with air at low pressure (or a partial vacuum, if you prefer)... cathode, anode, and anticathode.

Coolidge/Vacuum Tubes

Most x-ray tubes in use today are "filled" with a vacuum. This " entirely new variety " of x-ray tube was invented in 1913 by the American electrical engineer William Coolidge (1873–1975). In that same year Coolidge developed the technique for making fine wire out of tungsten (a notoriously non-ductile metal). Nearly every incandescent light bulb made after 1913 contains a tungsten filament made using Coolidge's process. When he was done working on light bulbs, he turned his attention to x-ray tubes. Guess what? Nearly every x-ray tube made after 1913 contains a tungsten filament made using the process used in light bulbs.

In a typical vacuum x-ray tube, electrons accelerated from a heated cathode toward a metal anode by a large potential difference. Changing the filament temperature changes the electron current — a hotter cathode releases more electrons than a cold one. This determines the intensity or "brightness" of the x-ray beam. Since one electron will produce one x-ray photon when it strikes the anode, more electrons flying through the tube means more x-ray photons emitted from the tube. The voltage across the tube determines the kinetic energy of the electrons when they strike the anode, which in turn determines the penetrating power of the x-ray photons — more energy per electron means more energy per x-ray photon and thus greater ability to plow through matter.

The cathode is a coiled filament of wire (usually tungsten) heated to around 2000 °C (white hot). It emits electrons through thermionic emission. In a sense, the electrons "boil" off the metal surface, but it's a weird kind of boiling since the electrons that leave are always replaced by new ones. If I put a pot of water on the stove at home, set it boiling and then leave the kitchen for an hour or two, by the time I get back there's a good chance the pot will be empty (and maybe even sizzling red hot). This does not happen with electrons in a cathode. The ones that leave are always replaced with new ones. If they didn't we'd wind up with a collection of positively charged ions (and eventually bare nuclei) that would surely fly apart due to their mutual repulsion. An x-ray tube is a circuit element. Current goes in one end and out the other and round and round the circuit.

The anode is a comparatively massive copper heat sink whose target face is cut diagonally and coated with some other metal (usually platinum). More than 99% of the kinetic energy imparted to the electrons is converted to heat on the anode. The remaining 1% is emitted as braking radiation (i.e., useful x-rays). This heat must be transferred or the target would melt. Coolidge's solution was to rotate the target using a small motor. This ensured that the hot spot never stayed in one place long enough to cause any lasting damage to the anode. (Some x-ray tubes are cooled with water.) The target is cut on a diagonal so that the emitted x-rays fly off the surface at an angle different from the incident electrons. A 45° cut makes the x-rays exit perpendicular to the axis of the tube. All the photographs of x-ray tubes on this page have their targets aligned at this angle. (The photo of a dental x-ray tube shown below left is a bit distorted, so the geometry isn't apparent.)

characteristic vs. bremsstrahlung (braking) spectra.

brems (braking/deceleration) + strahlung (radiation)

• In a cold pure metal (a), all electrons are below the Fermi energy level. Thermal energy allows electrons to form a space cloud in the vacuum (b), and application of an electric field allows the electrons to be collected on an anode; otherwise, an equilibrium is set up between the electrons inside and outside the metal. A tungsten wire is used in most x-ray tubes, electron microscopes and electron microprobes to take advantage of the high temperature for melting (3680 K) and evaporation. In a conventional x-ray tube, the wire is a coil approximately 1 cm by 1 mm, and the temperature is adjusted to minimize evaporation of W atoms which slowly contaminate the target. Unless an accelerating voltage is applied, there is no emitted current from a hot filament because of the formation of a space charge of electrons near the metal surface. The saturation current is measured by using the metal as a cathode of a vacuum tube and collecting the electrons on an anode which is sufficiently positive to dissipate the space charge. In a conventional x-ray tube, sufficient stability is obtained by regulating the filament voltage (for heating) and the accelerating voltage between cathode and anode.
• There are two (THREE?) principal mechanisms by which x-rays are produced. The first mechanism involves the rapid deceleration of a high speed electron as it enters the electric field of a nucleus. During this process the electron is deflected and emits a photon of x-radiation. This type of x-ray is often referred to as bremsstrahlung or "braking radiation". For a given source of electrons, a continuous spectrum of bremsstrahlung will be produced up to the maximum energy of the electrons.

X-rays are produced whenever fast moving electrons are decelerated, not just in x-ray tubes. Nearly all the naturally occurring x-ray sources are extraterrestrial. (No, that doesn't mean produced by alien creatures from outer space. It just means "beyond the Earth".) X-rays are produced when the solar wind is trapped by the Earth's magnetic field in the Van Allen Radiation Belts. Black holes are significant sources of x-rays in the universe. Matter falling into a black hole experiences an extreme acceleration caused by the intense field of the black hole. A single, isolated particle would fall in without releasing any radiation, but a stream of particles would as the particles would wind up crashing into each other on their way down the hole. Each inelastic collision experienced by a charged particle would result in the emission of a photon. Since these collisions are taking place at great speeds, the energies of the emitted photons in on the order of those found in the x-ray region of the electromagnetic spectrum. Inelastic collisions at even higher energies (greater than a million electronvolts) would generate gamma rays.

• The second mechanism by which x-rays are produced is through transitions of electrons between atomic orbits. Such transitions involve the movement of electrons from outer orbits to vacancies within inner orbits. In making such transitions, electrons emit photons of x-radiation with discrete energies given by the differences in energy states at the beginning and the end of the transition. Because such x-rays are distinctive for the particular element and transition, they are called characteristic x-rays .

The third mechanism is through synchrotron emission.

• Initially predicted in 1944 by Ivanenko and Pomeranschuk in Russia, it was, three years later, accidentally observed in a closed ring accelerator of the type of a synchrotron. It was long viewed as a "waste product", because synchrotron radiation is produced in the accelerators as a magnetic bremsstrahlung and undesirably limits the required final energy of the accelerators. Only several years later, in 1956, was synchrotron radiation specifically used in scientific investigations by Tomboulian and Hartmann.

Synchrotron radiation is emitted by charged particles traveling on a curved path (as would happen while moving through a magnetic field). Since the source of all electromagnetic radiation is the acceleration of charge, synchrotron radiation is an example electromagnetic radiation produced by centripetal acceleration (as opposed to bremsstrahlung, which is produced by tangential acceleration). The wavelength of this radiation is a function of the energy of the charged particles and the strength of the magnetic field bending the charged particles. The spectrum of the radiation is continuous and is characterized by its critical wavelength , which divides the spectrum into two parts with equal power (half the power radiated above the critical wavelength and half below).

The critical wavelength can be found using the equation below

which reduces to the following equation when the charged particles are electrons

Synchrotron radiation sources: rings, undulators, wigglers, National Synchrotron Light Source doesn't produce light as its primary form of electromagnetic radiation. Most research done at this facility uses the x-rays and vacuum ultraviolet produced by the electron beam.

• In 1945, the synchrotron was proposed as the latest accelerator for high-energy physics, designed to push particles, in this case electrons, to higher energies than could a cyclotron, the particle accelerator of the day. An accelerator takes stationary charged particles, such as electrons, and drives them to velocities near the speed of light. In being forced by magnets to travel around a circular storage ring, charged particles tangentially emit electromagnetic radiation and, consequently, lose energy. This energy is emitted in the form of light and is known as synchrotron radiation.

Synchrotron radiation is a nuisance in a particle accelerator as it sucks energy out of the particles being accelerated, but it makes an ideal source of high energy electromagnetic radiation. The beam produced is composed of nearly parallel rays ( collimated ) and is quite intense.

• Synchrotron radiation can be produced for hours, maybe even days if you were willing to pay the electric bils and had some reason to work around the clock. x-ray tubes can only operate for a few seconds or maybe minutes. Run them too long and they'll burn out just like a light bulb.
• Synchrotron radiation is "organized": the beam is highly polarized (most of the waves are oscillating in the same plane) and collimated (most of the waves are in the same direction). x-ray tubes produce "messy" radiation that is completely unpolarized and may be focused only with great difficulty. A synchrotron source is like an "x-ray laser", while an x-ray tube is like an "x-ray floodlight".
• Synchrotron radiation can be "shared". A large synchrotron might have upwards of 50 beam lines and run hundred if not thousands of experiments in one year. Synchrotron facilities are expensive to build, but pay for themselves in sheer volume of research.
• Wigglers or undulators (also known as insertion devices) produce synchrotron radiation that is considerably brighter than radiation from a bending magnet. The device causes electrons to follow a sinusoidal path instead of a curved one by establishing a series of magnetic fields that alternate in polarity and are perpendicular to the electrons' direction of travel. A wiggler enhances the brightness of the radiation produced by a given electron beam by a factor roughly equal to twice the number of full oscillations the beam undergoes. The deflections of the beam are smaller in an undulator than in a wiggler, and the radiation's brightness can, in theory, be increased by a factor about equal to the square of the number of oscillations, but only at discrete photon energies.

photon momentum

Max Planck discovered that phtons have energy.

E  =  hf

Albert Einstein discovered that energy and momentum are related.

E 2  =  p 2 c 2  +  m 2 c 4

Photons are massless, so this equation reduces to…

E  =  pc

Combine Planck and Einstein (their equations, not the men themselves)…

hf  =  pc

Solve for momentum…

Recall that…

If Planck and Einstein are correct, then photons have momentum too. What we need now is experimental evidence to support or refute this. (Don't worry. No one's going to refute this.)

compton effect

Arthur Compton (1892–1962) United States

shadowgraphs

Computed axial tomography (cat), x-ray scattering, x-ray diffraction.

Laue spots, Laue pattern. The German physicist Max von Laue (1879–1960) was the first to diffract X-rays on 12 April 1912 .

Bragg's law. The father-son team of of British physicists William Bragg (1862–1942) and Lawrence Bragg (1890–1971) derived this equation in 1913 .

n λ = 2 d  sin θ

DNA structure was determined from an x-ray diffraction image taken by Rosalind Franklin (1920–1958) England.

x-ray fluorescence

X-RAYS AND ENERGY

X-rays have much higher energy and much shorter wavelengths than ultraviolet light, and scientists usually refer to x-rays in terms of their energy rather than their wavelength. This is partially because x-rays have very small wavelengths, between 0.03 and 3 nanometers, so small that some x-rays are no bigger than a single atom of many elements.

DISCOVERY OF X-RAYS

X-rays were first observed and documented in 1895 by German scientist Wilhelm Conrad Roentgen. He discovered that firing streams of x-rays through arms and hands created detailed images of the bones inside. When you get an x-ray taken, x-ray sensitive film is put on one side of your body, and x-rays are shot through you. Because bones are dense and absorb more x-rays than skin does, shadows of the bones are left on the x-ray film while the skin appears transparent.

An x-ray image of teeth. Can you see the filling?

An X-ray photo of a one year old girl who swallowed a sewing pin. Can you find it?

Our Sun's radiation peaks in the visual range, but the Sun's corona is much hotter and radiates mostly x-rays. To study the corona, scientists use data collected by x-ray detectors on satellites in orbit around the Earth. Japan's Hinode spacecraft produced these x-ray images of the Sun that allow scientists to see and record the energy flows within the corona.

TEMPERATURE AND COMPOSITION

The physical temperature of an object determines the wavelength of the radiation it emits. The hotter the object, the shorter the wavelength of peak emission. X-rays come from objects that are millions of degrees Celsius—such as pulsars, galactic supernovae remnants, and the accretion disk of black holes.

From space, x-ray telescopes collect photons from a given region of the sky. The photons are directed onto the detector where they are absorbed, and the energy, time, and direction of individual photons are recorded. Such measurements can provide clues about the composition, temperature, and density of distant celestial environments. Due to the high energy and penetrating nature of x-rays, x-rays would not be reflected if they hit the mirror head on (much the same way that bullets slam into a wall). X-ray telescopes focus x-rays onto a detector using grazing incidence mirrors (just as bullets ricochet when they hit a wall at a grazing angle).

NASA's Mars Exploration Rover, Spirit, used x-rays to detect the spectral signatures of zinc and nickel in Martian rocks. The Alpha Proton X-Ray Spectrometer (APXS) instrument uses two techniques, one to determine structure and another to determine composition. Both of these techniques work best for heavier elements such as metals.

Since Earth's atmosphere blocks x-ray radiation, telescopes with x-ray detectors must be positioned above Earth's absorbing atmosphere. The supernova remnant Cassiopeia A (Cas A) was imaged by three of NASA's great observatories, and data from all three observatories were used to create the image shown below. Infrared data from the Spitzer Space Telescope are colored red, optical data from the Hubble Space Telescope are yellow, and x-ray data from the Chandra X-ray Observatory are green and blue.

The x-ray data reveal hot gases at about ten million degrees Celsius that were created when ejected material from the supernova smashed into surrounding gas and dust at speeds of about ten million miles per hour. By comparing infrared and x-ray images, astronomers are learning more about how relatively cool dust grains can coexist within the super-hot, x-ray producing gas.

EARTH'S AURORA IN X-RAYS

Solar storms on the Sun eject clouds of energetic particles toward Earth. These high-energy particles can be swept up by Earth's magnetosphere, creating geomagnetic storms that sometimes result in an aurora. The energetic charged particles from the Sun that cause an aurora also energize electrons in the Earth's magnetosphere. These electrons move along the Earth's magnetic field and eventually strike the Earth's ionosphere, causing the x-ray emissions. These x-rays are not dangerous to people on the Earth because they are absorbed by lower parts of the Earth's atmosphere. Below is an image of an x-ray aurora by the Polar Ionospheric X-ray Imaging Experiment (PIXIE) instrument aboard the Polar satellite.

Next: Gamma Rays

National Aeronautics and Space Administration, Science Mission Directorate. (2010). X-Rays. Retrieved [insert date - e.g. August 10, 2016] , from NASA Science website: http://science.nasa.gov/ems/11_xrays

Science Mission Directorate. "X-Rays" NASA Science . 2010. National Aeronautics and Space Administration. [insert date - e.g. 10 Aug. 2016] http://science.nasa.gov/ems/11_xrays

Discover More Topics From NASA

James Webb Space Telescope

Perseverance Rover

Parker Solar Probe

• History & Society
• Science & Tech
• Biographies
• Animals & Nature
• Geography & Travel
• Arts & Culture
• Games & Quizzes
• On This Day
• One Good Fact
• New Articles
• Lifestyles & Social Issues
• Philosophy & Religion
• Politics, Law & Government
• World History
• Health & Medicine
• Browse Biographies
• Birds, Reptiles & Other Vertebrates
• Bugs, Mollusks & Other Invertebrates
• Environment
• Fossils & Geologic Time
• Entertainment & Pop Culture
• Sports & Recreation
• Visual Arts
• Demystified
• Image Galleries
• Infographics
• Top Questions
• Britannica Kids
• Saving Earth
• Space Next 50
• Student Center
• Introduction
• Wave nature
• Particle nature
• Applications
• Production of X-rays
• Detection of X-rays

• What did Marie Curie accomplish?
• What awards did Marie Curie win?

Our editors will review what you’ve submitted and determine whether to revise the article.

• APS Physics - November 8, 1895: Roentgen's Discovery of X-Rays
• NHS - X-Ray
• Healthline - X-Ray
• Physics LibreTexts - X-rays
• Patient - X-ray
• National Institue of Biomedical Engineering and Bioengineering - X-rays
• National Center for Biotechnology Information - PubMed Central - The Scattering of X-Rays and Bragg's Law
• MedicineNet - X-Rays
• Portland State University - X-Rays and the Compton Effect
• Mayo Clinic - X-ray
• NASA - Chandra X-ray Observatory - The Science of X-ray Technology
• MSD Manual - Consumer Version - X-Rays
• Live Science - What are X-Rays?
• VeryWell Health - What is an X-Ray?
• Cleveland Clinic - X-Ray
• X-rays - Children's Encyclopedia (Ages 8-11)
• X ray - Student Encyclopedia (Ages 11 and up)
• Table Of Contents

Recent News

X-ray , electromagnetic radiation of extremely short wavelength and high frequency , with wavelengths ranging from about 10 −8 to 10 −12 metre and corresponding frequencies from about 10 16 to 10 20 hertz (Hz).

X-rays are commonly produced by accelerating (or decelerating) charged particles; examples include a beam of electrons striking a metal plate in an X-ray tube and a circulating beam of electrons in a synchrotron particle accelerator or storage ring . In addition, highly excited atoms can emit X-rays with discrete wavelengths characteristic of the energy level spacings in the atoms. The X-ray region of the electromagnetic spectrum falls far outside the range of visible wavelengths. However, the passage of X-rays through materials, including biological tissue , can be recorded with photographic films and other detectors. The analysis of X-ray images of the body is an extremely valuable medical diagnostic tool.

X-rays are a form of ionizing radiation—when interacting with matter, they are energetic enough to cause neutral atoms to eject electrons. Through this ionization process the energy of the X-rays is deposited in the matter. When passing through living tissue, X-rays can cause harmful biochemical changes in genes , chromosomes , and other cell components. The biological effects of ionizing radiation , which are complex and highly dependent on the length and intensity of exposure, are still under active study ( see radiation injury ). X-ray radiation therapies take advantage of these effects to combat the growth of malignant tumours.

X-rays were discovered in 1895 by German physicist Wilhelm Konrad Röntgen while investigating the effects of electron beams (then called cathode rays ) in electrical discharges through low-pressure gases. Röntgen uncovered a startling effect—namely, that a screen coated with a fluorescent material placed outside a discharge tube would glow even when it was shielded from the direct visible and ultraviolet light of the gaseous discharge. He deduced that an invisible radiation from the tube passed through the air and caused the screen to fluoresce . Röntgen was able to show that the radiation responsible for the fluorescence originated from the point where the electron beam struck the glass wall of the discharge tube. Opaque objects placed between the tube and the screen proved to be transparent to the new form of radiation; Röntgen dramatically demonstrated this by producing a photographic image of the bones of the human hand. His discovery of so-called Röntgen rays was met with worldwide scientific and popular excitement, and, along with the discoveries of radioactivity (1896) and the electron (1897), it ushered in the study of the atomic world and the era of modern physics .

Want to create or adapt books like this? Learn more about how Pressbooks supports open publishing practices.

2 From X-Ray Beam to Image Signal

In this chapter you will learn about the properties of the x-ray beam and how it’s properties change as it passes through matter. A satisfactory understanding of what takes place with the x-ray beam on its way to the image receptor is essential to understanding how changes in technical factors alter the beam’s properties and the resultant display image.

Learning Objectives

At the end of this chapter you should be able to:

• Differentiate between Source-Image Distance, Object-Image Distance and Source-Object Distance.
• Calculate SID, OID or SOD when given the two other factors.
• Differentiate between primary radiation and remnant radiation.
• Define attenuation.
• Explain the role electron binding energy plays in the interactions of x-rays with matter.
• Explain the interactions between x-rays and matter that occur at diagnostic ranges – Photoelectric effect and Compton Scattering
• Differentiate between secondary radiation and scatter.
• Discuss the factors that contribute to differential absorption.
• Define subject contrast.
• Differentiate between radiolucent and radiopaque tissues.

When a radiographic exposure is made, x-rays are produced from a small area of the anode called the focal spot. The x-rays diverge outwards from this area, travel in straight lines, and can be detected by a variety of image receptors. We frequently discuss portions of the x-ray beam in relation to the object that is being imaged. When the x-ray beam passes through the body, some photons in the beam will be absorbed by the tissues and organs, while other photons will pass through to the image receptor and form the basis for the creation of the display image.

Radiographic Geometry

Before learning about the factors that influence the accuracy of the radiographic image, it is important to learn some basic terminology associated with distances. For many radiographic studies, the patient lies flat on the x-ray table. The x-ray tube is positioned above them and the image receptor is placed in the Bucky tray below them. See Figure 2-1.

The distance between the x-ray tube and the patient can be measured, as can the distance between the patient and the image receptor. We can also measure the distance between the x-ray tube and the image receptor. These distances are important and have specific names.

Figure 2-1: Radiographic Distances

This diagram illustrates the various components of an x-ray beam with relation to an object (and objects internal to that object) and an image receptor.

Source-Image Distance

The distance between the x-ray tube and the image receptor is known and the source-image distance (SID) . This distance is called the source-image distance because it is the distance from the source of the x-ray beam, which is in the x-ray tube, to the spot where the image is formed, some type of image receptor. This distance can also be called the source-to-image receptor distance.

Object-Image Distance

The patient is inelegantly referred to as the object in these distance terms. However, referring to the patient as the object is not always precise. The patient’s body is a three-dimensional structure. If the radiographer is interested in producing a radiography of just one part inside the patient, the  that part becomes the object.

The distance between the part inside the patient’s body that is being radiographed and the place where the image is formed, which is usually at the image receptor, is known as the object-image distance (OID) .

Source-Object Distance

The last distance is the distance from the x-ray tube to the object. This distance is called the source-object distance (SOD) . This term is used less frequently in clinical situations than SID or OID, but it is used in several calculations.

Calculating the Distances

Activity 2A: Radiographic Distances

The X-ray Beam

When we discuss the x-ray beam, we typically think of the properties of the beam in terms of 2 concepts – quantity and quality.

Beam quantity refers to the number of x-rays in the beam. That makes sense, but another term that is used to refer to quantity is intensity . In normal speaking we think of intensity as synonymous with power. For the x-ray usage, that doesn’t work. Let us think about intensity of the x-ray beam by thinking about a flashlight. When someone shines a flashlight in your eyes, the brightness, or intensity of the flashlight is dependent on the number of light photons being given out.  If your battery is dying and your flashlight beam is less intense, then fewer light photons are being produced.

We can see x-ray quantity on the emission spectra of the x-ray beam. See Figure 2-2. When the quantity of x-rays in the beam changes, the amplitude of the graph moves up or down. In Figure 2-2, the x-ray beam represented by the pink line has a lower quantity or lower intensity than the x-ray beam represented by the blue line. Notice that the characteristic peaks move up or down as well as the overall brems curve. The primary factor that we use to control the quantity of the beam is mAs. Remember from our discussion of the production of x-rays that the number of electrons flowing through the filament controls the number of electrons that boil off and travel across the tube to make x-rays. We control the number of electrons flowing through the filament by the mAs we set. mAs stands for milliampere-seconds. This is because both mA (milliamps or the number of electrons flowing) and the time we let the electrons flow for (in seconds) both affect the number of x-rays created in the same way – double the mAs = double the # of x-rays.

Figure 2-2: Quantity of the X-ray Beam

The pink curve represents an x-ray beam with fewer x-rays in it. We would say that the pink x-ray beam has lower quantity or lower intensity than the beam represented by the blue curve.

The other idea we use to talk about the x-ray beam is its quality. The quality of an x-ray beam refers to its penetrating ability. We primarily control the quality of the x-ray beam using kVp. kVp is the potential difference or pull we use to accelerate the electrons from the filament to the anode. The faster we accelerate the electrons, the more kinetic energy they acquire. Electrons with more kinetic energy when they slam into the anode produce higher energy x-rays .

We can see the quality of the x-ray beam on the emission spectra as well. See Figure 2-3.  We see changes to the beam quality on the emission spectra as shifts to the left or to the right. When comparing the two x-ray beams represented in Figure 2-3, the blue beam extends further to the right and the top peak of the graph is also more to the right than the red beam. However, when we use kVp to change the quality of the beam, we also get a change in quantity. We see the change in quantity on the graph because the top of the blue line is higher than the red line.

Figure 2-3: Quality of the X-Ray Beam

Quality of the x-ray beam is represented by the left-right position of the graph. kVp is the primary controlling factor, but kVp affects both quality and quantity.

X-Ray Beam Before and After the Patient

The x-ray beam may be divided into two parts, the primary beam and the remnant beam. See Figure 2-4.

Figure 2-4: Primary and Remnant Radiation

The x-ray beam as it is emitted from the focal spot is referred to as primary radiation. When the x-ray beam passes through matter it undergoes attenuation. The portion of the x-ray beam that exits the patient and continues on to deposit energy in the image receptor is discussed as remnant or exit radiation.

Primary Radiation

The primary radiation is confined to the portion of the x-ray beam emitted from the focal spot of the x-ray tube. The radiation in the primary beam stays essentially the same from the point it is emitted from the x-ray tube until it encounters the object being examined. We also use the term incident x-rays or incident photons to refer to individual x-rays in the primary beam. Incident is a term that means “falling on or striking something,” so electrons crossing the tube and striking the anode are incident electrons and x-rays crossing the room and striking the patient are incident x-rays. The terms photon and quantum (or quanta in plural form) are referring to a single bundle of electromagnetic energy, i.e. a single x-ray. The terms photon and quantum also carry the idea of x-rays as a small particle. This is due to an interesting characteristic of x-rays- while they are electromagnetic waves, they sometimes behave as particles, even though they have no mass. This is the concept of wave-particle duality , and the reason x-rays can interact with matter at all. Wave-particle duality also applies to other high-energy forms of electromagnetic energy, such as gamma rays and cosmic rays.

Remnant Radiation

The remnant radiation (also sometimes referred to as exit radiation) is that portion of the primary beam that emerges from the examined object and interacts with the image receptor to record the radiographic image. But not all of the photons contained in the primary beam make it out of the object being examined. Some interact with the atoms in the organs and tissues inside the patient and are absorbed.  So, the remnant beam is what remains of the primary beam after the number of x-ray photons in the beam are diminished through absorption by the body tissues.

When using typical diagnostic exposure factors, only about 5% of the x-ray beam travels through the patient’s body without interacting and deposits energy in the image receptor to create the image.

Absorption vs. Attenuation

When x-rays pass through an object, some of the x-ray photons interact with atoms in the object and are no longer part of the x-ray beam. The reduction in the number of x-rays in the beam due to these interactions is called attenuation . See Figure 2-5. Attenuation occurs when an incident photon is either absorbed by the atoms of the tissue or scattered in another direction. There are five types of interactions that may occur between photons and the atoms in their path – Coherent or classical scattering, Compton Scattering, Photoelectric Absorption, Pair Production, and Photodisintegration. However, when dealing with x-ray energies that are typically found in diagnostic imaging x-ray beams, only two of these interactions routinely occur – Photoelectric Absorption and Compton Scattering.

Figure 2-5: Attenuation of the Beam

Electron Binding Energy

In order to understand the interactions of x-rays with matter we need to revisit the concept of electron shells and electron binding energy. You should recall from our discussion of the interactions that create x-rays inside the x-ray tube that atoms are surrounded by electrons that are positioned at specific distances from the nucleus called shells. These shells are labeled from the innermost shell outward starting with the letter K (See Figure 2-6). The positively charged nucleus exerts an electric force of attraction on the orbital electrons and the force is greater the closer the electron shell is to the nucleus. Because of the higher binding energies of  electrons closer to the nucleus, it takes more force to remove an electron from the K-shell than is does to remove an electron from one of the outer shells. Because it takes energy to fight the attractive force of the nucleus and allow the electron to maintain a position further from the nucleus, the electron itself has higher energy the further from the nucleus that it exists. The energy of the orbital electron is inversely proportional to the binding energy attracting it to the nucleus. This means that electrons in the O or P shells maintain higher energy levels than electrons in the K-shell.

While the tungsten atoms found in the x-ray tube have a very high atomic number and lots of electrons around them, most of the elements that make up the tissues in our bodies are much smaller. For example, a major element in our bones is calcium. Calcium has an atomic number of 20, and is therefore surrounded by 20 electrons. Because there are much fewer protons in a calcium nucleus the force holding electrons in each shell is much lower.

Figure 2-6: Bohr Model of a Calcium Atom

Calcium, which has an atomic number of 20, has 20 electrons surrounding the nucleus. Because the nucleus of calcium has a lower positive charge than tungsten, the binding energies holding the electrons in their shells are much lower.

Photoelectric Absorption

Photoelectric absorption, or the photoelectric effect, is essential to the formation of the radiographic image. The photoelectric effect takes place when the energy of the incident x-ray is just slightly greater than the binding energy holding the electron it collides with in place. In the process of photoelectric absorption, the energy of the x-ray photon is completely absorbed by an inner shell orbital electron. This added energy gives the electron the force needed to escape the binding energy of the nucleus and the electron is ejected from orbit. See Figure 2-7. We call the electron that is ejected from orbit during photoelectric interactions a photoelectron . The photoelectron will eventually collide with and ionize another nearby atom.

Figure 2-7: Photoelectric Effect

In photoelectric effect the energy of the incident x-ray is completely transferred to the inner shell electron and used to overcome the electron binding energy and propel the orbital electron away from the atom. The x-ray is completely absorbed in this process. The ejected electron is called a photoelectron.

Key Takeaways

In the photoelectric effect, because the photon is completely absorbed, the radiographic image will be left blank white in that area. The photoelectric effect is responsible for the production of high contrast in the image. Again, the photoelectric effect only occurs when the energy of the incoming photon is slightly greater than the binding energy that holds the orbital electron in place.

Secondary Radiation

Photoelectric effect creates an unstable atom by removing an inner shell electron. This hole in the inner shell is immediately filled by an electron from one of the outer shells (not necessarily the outermost shell, just further out than the shell with the vacancy). In order to transition to a shell closer to the nucleus, the electron must give up some of the energy that keeps it in a more distant electron shell.  The energy that is released is in the form of a characteristic x-ray. We call this secondary radiation because it is produced in the patient’s body and not in the x-ray tube. See Figure 2-8.

Figure 2-8: Secondary Radiation Production

Because the atoms in the human body have much lower atomic numbers than we see in the x-ray tube, the characteristic x-rays produced have much lower energy.  For example, calcium, which has one of the higher atomic numbers of the elements in the body, only has a k-shell binding energy of 4 keV. When an outer shell electron transitions to the k-shell, the secondary photon released will have an energy of just under 4 keV. Such low energy x-rays rarely make it out of the body, and only contribute to patient dose.  When we add substances with higher atomic numbers to the body, like the barium sulfate or water-soluable iodine contrast agents, the secondary radiation can reach the image receptor. For example, the k-shell binding energy of iodine is 33 keV, which is in the same energy range as the majority of the x-rays produced by brems interactions in the tube. Because secondary radiation doesn’t originate from the same focal spot, it does not represent the anatomy like the x-rays in the primary beam do. When secondary x-rays do reach the image receptor they create fog , additional radiation exposure that veils or hides the radiographic image produced by the remnant beam. This reduces the visibility of the image and is undesirable.

Compton Scatter

A process first described by Arthur Compton in 1922, the Compton effect involves a collision between an incident x-ray photon and an outer shell electron. Compton interactions occur when the energy of the incident x-ray photon is MUCH higher than the binding energy of the orbital electron it collides with. The incident photon collides with the outer shell electron and transfers some of its energy to the electron. The orbital electron gains enough energy to break the bonds holding it to the nucleus and propel it away from the atom. The remaining energy is emitted from the atom as an x-ray. See Figure 2-9. We call the electron that is dislodged from the atom is called a Compton or recoil electron , and the x-ray that is emitted a Compton scattered photon . Because an electron has been removed, the atom is now an ion. The recoil electron continues along its path until it binds with another nearby atom creating another ion.

Figure 2-9: Compton Interactions

In the Compton effect, the incident x-ray photon is partially absorbed by an outer shell electron.

When the remaining energy is emitted from the atom as a Compton x-ray, it has lower energy than the incident photon and is traveling in a different direction. That is why we refer to it as “ scatter .” The energy transfer can be described by the following mathematical equation:

Figure 2-10: Deflection Angles and Scattered Photon Energy

When the scattered photon has a small angle of deflection, it retains most of the energy of the incident photon. When the scattered photon’s angle of deflection is large, most of the energy from the incident photon is transferred to the electron and the scattered photon has much lower energy, much longer wavelength and much longer frequency.

The Compton scattered photon, or scatter for short, typically has high enough energy to reach the image receptor and cause destructive fogging of the image. Compton scattering is responsible for 99% of the fogging  photons that reach the image. The fog created by scattered photons, like the fog created by secondary radiation, does not correspond to anatomical structures and covers the image in radiation exposure that covers up, or veils, the anatomical signal and reduces image contrast. Many of the exposure choices we make as radiographers are aimed at reducing the number of scattered photons before they reach the image receptor. It is convenient that the strategies to control scatter are also effective and reducing or eliminating secondary radiation.

General Principles of X-Ray Interactions

X-rays interact with lots of different things, both inside and outside the x-ray tube. As imaging professionals, we are primarily interested in how they interact with our patient and the image receptors so we can optimize the display image and minimize patient dose.  Regardless of what structures we are imaging, the interactions are similar and are primarily affected by the average atomic number of the structure and the energy of the x-rays employed. Absorption is also influenced by the density of the tissue as well as its thickness.

The x-rays in the primary beam creates the image signal in the following ways (See Figure 2-11):

• X-rays that interact through the photoelectric effect contribute to the image by blocking the x-rays. This leaves areas of low exposure (white areas) on the image that represent areas of greater attenuation in the patient’s body.
• X-rays that pass through, or are transmitted through, the body deposit energy onto the image receptor. This leaves areas of greater exposure (black areas) on the image that represent areas of lower attenuation in the patient’s body.
• X-rays that are scattered inside the patient’s body, change direction but may still reach the image receptor. Scattered x-rays deposit energy into the image receptor and leave areas of greater exposure (black areas) on the image. These areas of exposure do not represent the patient’s anatomy but cover the light and dark areas evenly, making it more difficult to see the parts of the image that actually represent the anatomy. The scattered photons may be referred to as a type of noise because they hinder the visibility of the signal.

This demonstrates that, while Compton scatter and the photoelectric effect are important, it is the x-rays that are transmitted through the body without interacting that carry the image information to the image receptor.  Overall, the radiographic image results from the differences between the x-rays being absorbed by the photoelectric effect and the x-rays transmitted through the patient to the image receptor. The different  degrees of x-ray absorption in different tissues that creates image contrast is described as differential absorption .

Figure 2-11: Differential Absorption and the Radiographic Image

The x-rays in the primary beam creates the image signal when passing through the patient by:

• Photoelectrically absorbed x-rays leave areas of low exposure (white areas) on the image that represent areas of greater attenuation in the patient’s body.
• X-rays that pass through the body leave areas of greater exposure (black areas) on the image that represents areas of lower attenuation in the patient’s body.
• Scattered x-rays  leave areas of greater exposure (black areas) on the image but contribute no useful information.

Differential Absorption

When x-rays pass through the body, each type of tissue absorbs the radiation in its own way. We refer to this difference in the number and type of radiation interactions between two tissues as differential absorption . The amount of radiation the tissue absorbs corresponds to the atomic number of the elements composing the tissue as well as the thickness of the tissue and its density. The atomic number of the elements that compose the tissue causes bone to absorb more x-rays than muscle or fat. Differential absorption is a key factor determining image contrast. Contrast is an idea that we will discuss throughout this book. Contrast is the number of black, white and gray tones seen in an image. It can also be defined as the difference in brightness between two adjacent areas of the image. Differential absorption and the subject contrast it creates is the primary reason we can tell the difference between different tissues within the same patients.

We also use differences in atomic number to make some structures more apparent by using contrast media. Contrast agents, which contain either barium (atomic number 56) or iodine (atomic number 53) enhance the visibility of certain anatomical structures because they contain atoms with higher atomic numbers than the surrounding tissues. The higher atomic number results in greater photoelectric interactions with the beam.

When we are comparing radiation absorption between patients, the tissue thickness and density comes into play. For example, muscle tissue is generally composed of the same elements for everyone, but a bodybuilder has thicker muscles and will absorb more radiation than the same body part of an individual who does not work out. Similarly, the measured thickness of the bodybuilder’s arm may be the same as the thickness of an obese person’s arm, but the muscle tissue is more dense than fat and, again, absorbs more x-rays. In this example we see that x-ray absorption cannot be predicted by thickness alone. We must take into consideration the individual’s physical state and the possible presence or absence of elements with higher atomic numbers. The absorption of x-rays by various tissue components introduces the idea of subject contrast, which is of major importance to the creation of the radiographic image.

Subject Contrast

We refer to the differences in x-ray absorption in different tissues as subject contrast . We use some special terms to talk about the relative differences in x-ray absorption. Tissues that produce a high amount of photoelectric interactions appear white on the image and are referred to as radiopaque . Bone is an excellent example of a radiopaque tissue. Tissues that allow the x-rays to pass through without interacting appear dark or black on the image and are referred to as radiolucent . Lung tissue is an excellent example of radiolucent tissue.

The combination of radiopaque and radiolucent structures lets us understand the patient’s anatomy. See Figure 2-9. In a typical chest x-ray we see several bony radiopaque structures including the spine, clavicle and ribs, but we also see the heart shadow as radiopaque because it absorbs more radiation than the surrounding lung tissue. The thoracic spine is more radiopaque than the heart, because we can see the individual vertebrae through the outline of the heart.  We also see several radiolucent structures on a typical chest x-ray. These structures are typically filled with air. Because air is a gas, the atoms are not dense at all and do not block x-rays well. In figure 2-12, we see the lungs as radiolucent, but we also see the trachea filled with air, where it overlies the top portion of the thoracic spine.  The air above the patient’s shoulders likewise appears dark on the image because there was nothing in the way of the x-ray beam for the photons to interact with.

Figure 2-12: Subject Contrast in a Chest X-Ray

On a typical chest x-ray the spine, clavicles, ribs and heart appear radiopaque. The lungs and trachea appear radiolucent.

Activity 2B: The X-Ray Beam and its Interactions

Test your understanding of concepts related to the x-ray beam and its interactions with matter on this crossword puzzle.

LAB Exercise 2: Discovering Subject Contrast

Radiographic distances include the distance between the x-ray tube and the image receptor – SID, the distance between the object being radiographed and the image receptor – OID, and the distance between the x-ray tube and the object being radiographed – SOD.  These distances can be calculated if you know 2 of the 3 factors.

Attenuation is the reduction in the number of x-ray photons in the beam as it travels through an object. Attenuation happens when x-rays in the beam interact with the matter it is passing through.

The two interactions between x-rays and matter that occur at diagnostic levels are Compton scattering and the photoelectric effect.

• In Compton scatter the incident x-ray photon interacts with an outer shell electron, ionizing the atom and redirecting the x-ray that has lost some of its energy. The electron ejected from the atom in a Compton interaction is called a recoil electron.
• In photoelectric effect, the incident x-ray photon interacts with an inner shell electron where it is completely absorbed in the process of ionizing the atom. The electron ejected from the atom in a photoelectric interaction is called a photoelectron.

Differential absorption is the differences seen the degree of absorption of the x-ray beam between different tissues. These differences in x-ray absorption are responsible for creating the radiographic image.  Differential absorption occurs because of differences in:

• The atomic number of atoms in the tissue
• The x-ray energy
• The mass density of the tissue
• The thickness of the part

Contrast agents enhance the visibility of certain anatomical structures because they contain atoms with higher atomic numbers, and have higher mass density than the surrounding tissues.

We use variations in x-ray energy, primarily controlled by kVp, to vary the differential absorption, and subsequently the contrast of our images.

While we should consider the thickness of the part when selecting technical factors, we must also consider the make-up of the area in terms of tissue type and mass density.

The x-ray that is in the exact center of the x-ray beam.

The portion of the imaging chain that captures the x-ray signal for conversion into a visible image. Abbreviated "IR".

The object-image distance; the distance between the image receptor and the part inside the patient that is being examined.

The source-image distance; the distance from the x-ray tube to the image receptor.

Digital Radiographic Exposure: Principles & Practice Copyright © 2022 by Carla M. Allen. All Rights Reserved.

Share This Book

National Aeronautics and Space Administration

Goddard space flight center, imagine the universe.

• Astronomer's Toolbox
• Cosmic Objects
• Big Questions
• Featured Science
• Observatories
• Scientist Profiles
• You Be the Astrophysicist
• The Cosmic Distance Scale
• Lesson Plans
• Ask an Astrophysicist
• Other Resources
• News #include virtual="/news/newsNav.html"

Educators' Corner

Background: introduction to spectroscopy, introduction to spectroscopy.

The Electromagnetic Spectrum

Line emission, different graphical representations of spectra, spectra and astronomy, continuum emission.

What Are Some Examples of Continuum Emission?

Reference URLs:

• Project Leader: Dr. Barbara Mattson
• Web Curator: J.D. Myers
• Responsible NASA Official : Dr. Andy Ptak
• Privacy Policy & Important Notices
• Page Last Updated: 14-Oct-2021

Junior doctors & students menu

Chest X-rays for Medical Students

A superb guide to CXRs for students

Abdominal X-rays for Medical Students

A superb guide to AXRs for students

Radiology B a s i c s

Ultrasound and CT e-learning for junior doctors and students

Join 10k+ newsletter subscribers

Please note: Your email address will never be shared with any 3rd parties. It will only be used for Radiology Cafe communications. Emails are sent less than once a month on average. Read our Privacy policy for more details.

How to present an X-ray

A short guide to help you improve your presentation of radiographsin an exam situation or on the ward

The purpose of this guide isn’t to teach you what you’re going to see, but how you should say it!   Guides to help you get through looking at images systematically can be found elsewhere; our aim is to improve your presentation skills.

Opening line

Imagine you’re sat in an OSCE.  You are presented with a radiograph and invited to comment on it.  How might you start presenting?  Just as you wouldn’t jump straight to the heart sounds when presenting a cardiovascular exam, so you shouldn’t jump straight to the interesting bit of the image.

An opening line of:

“frontal chest radiograph of an adult patient”

…can virtually never be wrong.  This is often enough as an opening line, but you could expand if you can with the details you’re given – especially if you are given some of the patient’s history. For example:

“frontal chest radiograph of an adult male patient with shortness of breath”

For the first radiograph you see, it’s good to briefly demonstrate you know what an adequate radiograph looks like.  DO NOT SPENT LONG DOING THIS (10 seconds at most).  Some medical schools do not award any marks for talking about the adequacy of a radiograph so whatever you do, do not spend much time on this!  The interesting bit is describing the pathology on the radiograph itself.  On your second or subsequent radiograph during a station, if the image is clearly adequate, simply stick with your single opening line and proceed to the lungs.

There are broadly 3 categories of adequacy to assess:

• Rotation: are the clavicles equidistant from the spinous processes?
• Penetration: can you see the spine through the heart?
• Expansion: The anterior aspect of the 6 th rib (the sloping ones) should meet the diaphragm near the mid-clavicular line. You should be able to see the apices and costophrenic angles.

So you might introduce the above radiograph as follows:

“frontal chest radiograph of an adult female patient. It is not rotated, with good penetration and good expansion of the lungs, although the tips of the costophrenic angles are not included on this radiograph”

As a note of caution, some medical schools do not award any marks for talking about the adequacy of a radiograph so whatever you do, do not spend much time on this!  The interesting bit is describing the pathology on the radiograph itself.  On your second or subsequent radiograph during a station, if the image is clearly adequate, simply stick with your single opening line and proceed to the lungs.

Main abnormality

So you’ve made it to the lungs and something looks wrong.  This almost always means there is opacity (whiteness) where there shouldn’t be.  Describe this in terms of appearance and location.

Is the opacity:

• Patchy – hard to draw round properly, heterogeneous (e.g. pulmonary oedema)

• Dense – very white so that you can barely see structures through it (e.g. pleural effusion)

• Rounded – discrete, round(ish) in shape (e.g. a mass lesion in the lung)

Then ask yourself, where is the lesion located?

The lung lobes overlap significantly on a frontal chest radiograph, so it is hard to place an abnormality in a specific lobe when viewed only from the front.  A solution to this is to split the lung into zones – upper zone is above the heart, mid-zone is level with the top half of the heart, and lower zone is below that.  It is also worth looking under the diaphragms and behind the heart for subtle, ‘hidden’ lesions.

Ok, so now think how you might describe the abnormality in the radiograph we showed you at the start.

“there is a solitary rounded opacity in the left upper zone”

Review areas

To complete your assessment of a radiograph, particularly if it’s your first of a session, it’s worth quickly summarising the rest of the radiograph to show you have assessed everything.  If the abnormality is in the lungs, you might also comment on:

• The heart: is it enlarged?
• The diaphragm: can you see any free gas?
• The bones: are there any fractures?

So for our radiograph:

“the heart is not enlarged, there is no free air below the diaphragm, and there are no fractures identified”

Put everything above together and you’ll have had 30 seconds or so to think up a differential diagnosis to put the icing on your presentation:

“this could represent a primary lung malignancy, but might also be a metastasis”

Finally, finish by saying what you would do next:

“I would like to request a CT chest to further evaluate the lesion and look for evidence of malignancy”

So that’s it!

Stick to these basic rules and you will go a long way in your exams and when presenting on the ward.  Good luck!

• Christopher Clarke
• Last updated: 10 October 2021

Interview & application

Radiology trainees, junior doctors & students.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Biomed Res Int

Modern Diagnostic Imaging Technique Applications and Risk Factors in the Medical Field: A Review

Shah hussain.

1 Department of Biochemistry, Faculty of Biological Sciences, Quaid-i-Azam University Islamabad, Pakistan

Iqra Mubeen

Niamat ullah.

2 Department of Pharmacy, Faculty of Biological Sciences, University of Malakand, Pakistan

Syed Shahab Ud Din Shah

Bakhtawar abduljalil khan.

3 Women's Wellness and Research Center, Hamad Medical Corporation, Doha, Qatar

Muhammad Zahoor

4 Department of Biochemistry, University of Malakand, Chakdara, Dir Lower, KPK, Pakistan

5 Department of Pharmacognosy, College of Pharmacy, King Saud University, Riyadh, Saudi Arabia

Farhat Ali Khan

6 Department of Pharmacy, Shaheed Benazir Bhutto University, Sheringal, Dir Upper, KPK, Pakistan

Mujeeb A. Sultan

7 Department of Pharmacy, Faculty of Medical Sciences, Aljanad University for Science and Technology, Taiz, Yemen

Associated Data

This is a review article. All data are taken from published research papers and available online.

Medical imaging is the process of visual representation of different tissues and organs of the human body to monitor the normal and abnormal anatomy and physiology of the body. There are many medical imaging techniques used for this purpose such as X-ray, computed tomography (CT), positron emission tomography (PET), magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT), digital mammography, and diagnostic sonography. These advanced medical imaging techniques have many applications in the diagnosis of myocardial diseases, cancer of different tissues, neurological disorders, congenital heart disease, abdominal illnesses, complex bone fractures, and other serious medical conditions. There are benefits as well as some risks to every imaging technique. There are some steps for minimizing the radiation exposure risks from imaging techniques. Advance medical imaging modalities such as PET/CT hybrid, three-dimensional ultrasound computed tomography (3D USCT), and simultaneous PET/MRI give high resolution, better reliability, and safety to diagnose, treat, and manage complex patient abnormalities. These techniques ensure the production of new accurate imaging tools with improving resolution, sensitivity, and specificity. In the future, with mounting innovations and advancements in technology systems, the medical diagnostic field will become a field of regular measurement of various complex diseases and will provide healthcare solutions.

1. Introduction

Medical imaging is the process of visual representation of the structure and function of different tissues and organs of the human body for clinical purposes and medical science for detailed study of normal and abnormal anatomy and physiology of the body. Medical imaging techniques are used to show internal structures under the skin and bones, as well as to diagnose abnormalities and treat diseases [ 1 ]. Medical imaging has changed into healthcare science. It is an important part of biological imaging and includes radiology which uses the imaging technologies like X-ray radiography, X-ray computed tomography (CT), endoscopy, magnetic resonance imaging (MRI), magnetic resonance spectroscopy (MRS), positron emission tomography (PET), thermography, medical photography, electrical source imaging (ESI), digital mammography, tactile imaging, magnetic source imaging (MSI), medical optical imaging, single-photon emission computed tomography (SPECT), and ultrasonic and electrical impedance tomography (EIT) [ 2 ].

Imaging technologies play a vital role in the diagnosis of abnormalities and therapy, the refined process of visual representation which contributes to medical personnel access to awareness about their patient's situation [ 3 , 4 ]. Electroencephalography (EEG), magnetoencephalography (MEG), and electrocardiography (ECG) are recording and measurement techniques that are not responsible to produce images, but these represent the data as a parameter graph vs. time or maps which shows the susceptible information with less accuracy. Therefore, these technologies can be said to form medical imaging on a limited scale. Worldwide, up until 2010, approximately 5 billion medical imaging techniques studies have been shown [ 5 ].

In the United States, approximately 50% of total ionizing radiation exposure is composed of radiation exposure from medical imaging [ 6 ]. Medical imaging technologies are used to measure illnesses, manage, treat, and prevent. Nowadays, imaging techniques have become a necessary tool to diagnose almost all major types of medical abnormalities and illnesses, such as trauma disease, many types of cancer diseases, cardiovascular diseases, neurological disorders, and many other medical conditions. Medical imaging techniques are used by highly trained technicians like medical specialists, from oncologists to internists [ 1 ].

Medical imaging technologies are mostly used for medical diagnoses. Medical diagnosis is the process of identification of patient disease and its symptoms. The medical diagnosis gives the information about the disease or condition needed for treatment that is collected from patient history and physical checkups or surveys. Due to no specificity of the many signs and symptoms of a disorder, its diagnosis becomes a challenging phase in medical science. For example, the case of erythema (redness of the skin) gives a sign of many diseases. Thus, there is a need for different diagnostic procedures the determination the causes of different diseases and their cure or prevention [ 7 ].

Historically, the first medical diagnosis composed by humans was dependent upon the observation of ancient doctors with their eyes, ears, and sometimes examination of human specimens. For example, the oldest methods were used to test on body fluids like urine and saliva (before 400 B.C). In ancient Egypt and Mesopotamia, doctors were able to measure the problem of the digestive tract, blood circulation, heartbeat, spleen, liver menstrual problems, etc. But unfortunately, medicine for curing diseases was only for wealthy and royal people.

At around 300 B.C., the use of the mind and senses as diagnostic tools was promoted by Hippocrates. He got a reputation as the “Father of Medicine.” Hippocrates supported a diagnostic protocol by testing the patient urine, observing the skin color, and listening to the lungs and other outward appearances. The link between disease and heredity also had been recorded by them [ 8 ]. In the Islamic world, Abu al-Qasim al-Zahrawi (Arabic physician) provided the first report on a hereditary genetic disease referred as hemophilia. In this report, he wrote about a family of Andalusia, whose males died due to hemophilia [ 9 ]. In the Middle Ages, many different techniques were used by physicians to detect the causes of imbalance function of the body. Uroscopy was the most common method of diagnosis. The patient's urine was collected in a special type of flask known as “Matula.” Urine was checked on the basis of color, smell, density, and presence of precipitate [ 10 ]. The viscosity and color of the blood were also examined by physicians to detect chronic or acute diseases [ 11 ]. The pulse rate, power, and tempo of a patient's artery were observed by physicians through a technique known as palpation [ 12 ]. In Middle Ages, physicians were also used to combining the study of medicine and zodiac signs [ 13 ].

In the 19 th century, X-rays and microscopes were the diagnostic tools that helped to diagnose and treat illnesses. At the beginning of the 19 th century, medical doctors diagnosed diseases by the examination of symptoms and signs. By the 1850s, many diagnostic tools such as ophthalmoscopes, stethoscopes, and laryngoscopes lead to evoke the medical doctors with the sensory power to develop other novel methods and techniques for diagnosing different illnesses. And in this way, a series of diagnostic tools including chemical tests, bacteriological tests, microscopic tests, X-ray tests, and many other medical tests were generated [ 8 ].

Medical imaging techniques are developed after the discovery of X-rays. In November 1895, Wilhelm Conrad Roentgen discovered X-ray. He got the Nobel Prize in 1901 for his discovery. Radiologists gave names to X-ray basis as “X-rays” or “plane film” used for diagnosing bone fractures and chest abnormalities. Fluoroscopy was developed due to a more powerful beam of X-ray for diagnosing the patient abnormalities. In 1920s, radiologists started giving information about various diseases like cancer of the esophagus, ulcers, and stomach. Fluoroscopy is now converted into computed tomography (CT).

Today CT scan is commonly used to diagnose many diseases. The mammography technique also uses an X-ray beam, to generate high-resolution breast images, monitoring breast cancer. In the 1940s, the X-ray tomography technique was developed, looking for a desired part of the tissue. In this technique, the whole process was accomplished by rotating the tube of X-ray focus on part of the tissue. Today, tomography is replaced with advanced imaging techniques such as CT scanning or computerized axial tomography (CAT) scanning. X-ray is also a source of a technique known as “angiography,” which is used to obtain images of blood vessels. In 1950s, diagnostic imaging tests along with nuclear medicine were started. Radioactive compounds are used as X-ray sources rather than X-ray tubes. Radioactive compounds produce gamma rays. They are joined with other complexes that are an essential part of the disease analysis to study a certain illness. For an instant, technetium 99m is combined with methylene diphosphonate, which is absorbed by bone tumors. In this way, breast or lung cancer spread to other body parts such as bone can be detected from this type of nuclear bone scan technique [ 14 ].

2. Advance Modalities in Medical Imaging

Many advanced techniques are developed and can be explained with their principle of work, application in medical labs, and development in imaging techniques. Computed tomography (CT), positron emission tomography (PET), magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT), digital mammography, and sonography are included in advanced medical imaging techniques. These are all mentioned below to understand their advantages and applications in the diagnosis, management, and treatment of different diseases such as cardiovascular disease, cancer, neurological illnesses, and trauma. These techniques are readily used by clinicians because through images, they can easily choose how to manage diseases.

3. Computed Tomography (CT)

In the 1969s, Hounsfield invented the first CT-scanner prototype [ 15 ]. Computed tomography is also known as X-ray CT. A CT scan is used by radiologists, biologists, archaeologists, and many other scientists to generate cross-sectional images of different scanned objects. A modern CT scanner system is shown in Figure 1 . In the medical field, technicians use CT scanners, machines to produce the images that lead to diagnosing the abnormalities and other therapeutic measurements. In this technology, X-rays are produced from different angles that are eventually processed by computers to create tomographic images. This computer-based technology has been greatly improved, developing reconstructed images with high revolution [ 16 ]. In the pharmaceutical industry, it has been used to study and improve the medicine manufacturing process to generate good quality products [ 17 ].

CT scanner.

Positron emission tomography (PET) and single-photon emission computed tomography (SPECT) are the types of CT scan. An X-ray generator is used to generate the X-rays that rotate nearby the object to be scanned. X-rays are detected by an X-ray detector located on the opposite side of the source of X-rays. A sonogram is obtained, which is a visual representation of raw form data. This scanned data is processed in the form of tomographic reconstruction that leads to generating a series of cross-sectional photos. CT scan is performed by special individuals called radiology technologists or radiographers. Over the last two decades, CT is used largely in many clinical labs in different countries [ 18 ]. According to an estimated study, almost a 72 million scans were achieved in the US in 2007 and 80 million scans in 2015 [ 19 ].

CT is an effective technique for monitoring various types of cancers such as cancer of the bladder, kidneys, skeleton, neck, and head and for diagnosing infection [ 20 – 22 ]. CT also identifies distant metastases to the lungs, skeleton, liver, and brain. CT has made a high impact on the brain and lungs [ 23 , 24 ]. CT scan is the best method than other techniques in detection as well as recording modifications in tumor mass during treatment [ 25 ]. It may show a bloated belly with enlarged lymph nodes in patients with bronchus carcinoma. In this way, CT scan help in performing before surgery [ 26 ]. Another major application of CT scans is the detection of heart diseases like myocardial disease, congenital heart disease, and coronary artery bypass grafts [ 27 ]. Gastroenterologists mostly use computed tomography for the analysis of the liver or pancreas of patients. Tumors of size 1.5-2.0 cm in diameter can be detected by CT scan. Furthermore, biliary obstruction caused by lesions can also be monitored by this technology [ 28 ]. One of the rewarding roles of technology is to study suspected intra-abdominal abnormalities with 95% accuracy, and treatment decisions can be easily made [ 29 ].

A big drawback of computed tomography is that large masses within the gastrointestinal tract may not be visible during the abdominal investigation. There is also no finding of some of the mucosal abnormalities by it. CT scan is highly useful to manage abdominal disorders such as carcinoma of the stomach, esophagus, and rectum more accurately as compared to other modalities [ 30 , 31 ] The middle column of the spine can be visualized by using the computed tomography technique during dislocation type of fractures in many thoracolumbar fractures. CT also detects lesions and provides nonsurgical management of some disorders, for example, unstable burst damages [ 31 ].

All spinal injuries are unstable and known as translational injuries. Before surgery of such patients, complete information about the site of ligament discontinuity of respective vertebrae is provided by computed tomography. It also gives the prediction of whether Harrington-rod stabilization is possible or not. CT scan can provide detailed evidence of distraction injuries and fractures. For example, flexion-distraction injury between the 11 th and 12 th thoracic vertebrae and spinal injury between the 2 nd and 3 rd lumbar vertebrae have been scanned by computed tomography scan as shown in Figures ​ Figures2 2 and ​ and3, 3 , respectively [ 32 ].

Distraction injury scanned by CT scan (showing damage occurrence at the 11 th and 12 th thoracic vertebrae).

An axial scan of a spinal injury by computed tomography (CT) at the 2 nd and 3 rd lumbar vertebrae.

Advanced CT bone imaging techniques include volumetric quantitative CT (QCT), high-resolution CT (CT), and micro-CT. High-resolution CT and high-resolution MR are generally used in vivo ; micro-CT and micro-MR are usually used in vitro systems. These advanced modalities are used for bone imaging to investigate bone diseases especially osteoporosis and bone cancer. In osteoporosis, disorder advanced CT bone imaging provides information about bone mineral density (BMD), bone strength, a risk factor for osteoporosis, and recovery factors after medication or bone therapy.

Dual-energy X-ray absorptiometry (DXA) and volumetric QCT are the quantitative methods used for weighing the macrostructure of suspected bone. High-resolution CT and micro-CT methods are applicable for measuring the microstructure of trabecular bone without any invasiveness or destructiveness. CT and MRI have been used to obtain bone structure. However, the CT field has been more developed as compared to other techniques, because there are more advantages of CT-based modalities; for example, QCT generates three-dimensional (3D) images in such a way that trabecular and cortical bone can be distinctly measured. vQCT technology is quicker than MRI [ 33 ].

4. Volumetric Quantitative CT (vQCT)

Initially, QCT has been used to measure trabecular BMD of the forearm and lumber midvertebrae through a particular transverse CT slice. The measurement of BMD is a static property of the advanced spiral QCT [ 34 ]. Trabecular bone in the spine and cortical bone in the hip may be indicated by this technology to estimate the fracture risk [ 35 ].

For the improvement of the 3D structure of the cortex, almost 0.5 mm isotropic spatial resolution is required, but still, almost 1.5 to 2 mm resolution is provided by QCT which is not adequate to make accurate images. This is a drawback of QCT. In general, the measurement of accurate cortical thickness for the femur is easier than the thickness in the spine, especially in aged people. Researchers have shown that women grow faster not only with small vertebrae as well as reduced bone mass but also with a slow rate of increase in cross-sectional area as compared to men [ 35 ]. QCT is a CT imaging technique that can provide information about bone density. For example, a QCT scan of the femur for the measurement of macrostructure and bone mineral density (BMD) has been shown in Figure 4(a) [ 33 ].

(a) Femora undergo vQCT to determine BMD and macrostructure. (b) Ultradistal forearm undergoes CT to measure the structure of the trabecular complex network and its texture.

5. High-Resolution CT (hrCT)

High-resolution CT is a modern CT scanner that usually requires a high radiation dose but produces high-resolution images of bone such as forearm bone submitted to CT to determine the trabecular and cortical network and texture as shown in Figure 4(b) [ 33 ]. According to many different cross-sectional studies, CT gave better imaging results in distinguishing fractured vertebral trabecular structures from nonfractured structures as compared to DXA measurements of BMD [ 36 ].

6. Micro-CT ( μ CT)

Micro-CT with 1-100  μ m spatial resolution is typically known as microscopy. Micro-CT has abilities to replace the standing techniques used in in vivo measurements in rats and mice like animals. Initially, the micro-CT technique used synchrotron radiations to obtain ultra-high-resolution applications [ 37 ].

Mostly, now, the convenient method of X-ray tube-based micro-CT is used in university-based research laboratories and special clinical centers. To make 3D structures of bone, some special software (for example, FEM) is attached with a micro-CT scanner. Finite element modeling (FEM) is a software mostly used in engineering. Its goal is to help information of 3D structures of bone for analysis of fractured bone part structures from nonfractured bone structures. Currently, structural models are generated by volumetric QCT, and computer-based programs give the element elastic properties from the bone density at the site of elements [ 38 ].

Arlot and coworkers determined the 3D microstructure of bone of postmenopausal women with osteoporotic disease; they have accomplished treatment with proper strontium ranelate therapy for 36 months [ 39 ]. Researchers investigate these 3D micro-bone structures as shown in Figure 5 [ 33 ]. Over the past two decades, considerable development occurred in imaging technologies for osteoporosis bone disease analysis. Despite the development in these technologies, there are many challenges for bone imaging, such as the sample size, spatial resolution, complexity, radiation exposure, time, and cost. Finally, there is still a requirement of high accuracy, availability, reproducibility, and proper monitoring procedure for better bone imaging [ 33 ].

Microstructure of transiliac bone biopsies is determined to undergo 3D micro-CT of two postmenopausal women who have accomplished strontium ranelate therapy for 36 months: (a) strontium ranelate therapy; (b) placebo.

CT scan for lung disease is highly used in present days. For example, a low-radiation helical chest CT scan is used to investigate lung cancer (bronchopulmonary cancer) [ 40 , 41 ]. Another lung disease is the most common type of progressive idiopathic interstitial pneumonia mostly in adults known as idiopathic pulmonary fibrosis (IPF). In IPF patients, CT-based methods include density histogram analysis, CT scan of whole lungs, and density mask technique, and other structural or texture classification methods are greatly used to examine the pulmonary function, lung disease progression, and mortality. For example, lung images of a 73-year-old male IPF patient have been taken by CT as shown in Figure 6 . These methods have the property of time efficiency, availability, and reproducibility. Still, there are many issues interrelated to computer-based CT in IPF disease analysis. But it is promising by scientists to develop advanced CT imaging techniques that must play a vital role in the future to manage lung diseases as well as other abdominal diseases [ 42 , 43 ].

Images derived using a system known as GHNC (Gaussian Histogram Normalized Correlation). Lung images of a 73-year-old male IPF patient have been taken by computed tomography (CT). Light blue and yellow color, fibrosis; dark blue color, emphysema; pink color, normal; and light green color is indicated as ground-glass opacity.

There are many possible reasons for the usage of CT scans, for example, to determine or investigate the acute stroke in the patient's head. CT is applicable to establish the diagnosis, investigate the type of stroke, respond to surgery, and finally manage the disease [ 44 ]. CT scan of the head is also responsible for the investigation of dementia disease. Accurate diagnosis is directly related to proper management of symptoms and signs of dementia. Patients with treatable lesions can also be identified by computed tomography [ 45 ]. Abdominal computed tomography is a new technology for identifying fungal infection known as disseminated fungal infection (DFI) in pediatric cancer patients. Currently, abdominal CT is greatly applicable for the diagnosis and management of DFI in cancer patients [ 46 ]. During the past few years, the usage of CT scan has become a national trend in emergency departments, especially in the US. Computed tomography plays an expanding role in diagnosing acute and chronic diseases as well as life-threatening diseases such as stroke, head injury, major trauma, heart disease, abdominal pain, pulmonary embolism, severe chest pain, and renal abnormalities [ 47 ].

7. 3D Ultrasound Computed Tomography (3D USCT)

3D USCT is a promising technology for imaging breast cancer. Simultaneous recording of reproducible reflection, speed of sound volume, fast data collection, attenuation, and high image quality production are all the main advantages of the USCT system. 3D USCT system is a full potential device used for clinical purposes. Only in 4 minutes, the full volume of breast can be picked up [ 48 ].

8. Risks of Computed Tomography

Computed tomography risks are small but if these small risk s are produced by million numbers of scans, they may drive into serious public health concerns in the future, especially from pediatric CT. The risk of cancer due to computed tomography scanning is increasing [ 49 ]. Children are of specific concern due to the sensitivity of radiation-induced cancer as compared to adults. According to a study, the risks of leukemia and brain tumors are mostly revealed after exposure to radiation from CT scans [ 50 ].

According to the authors of a recent report “long-term risks from CT scans directly would require very large-scale studies with lifelong follow-up.” The author gave this statement after the study date on CT scan exposure leading to the risk of future cancer [ 51 ]. Radiologists should be a source of discussion earlier to perform imaging technologies that contain high doses of radiation. They should also know about the risk factors of imaging technologies that can cause more adverse effects than recovery. Both families and patients should also raise the question/answers about the benefits and risks of CT scans [ 52 ].

9. Positron Emission Tomography (PET)

Positron emission tomography is a nuclear medicine functional technique that is used to display the total concentration of radioactive labeled elements in the body with clear images. It has the potential to diagnose biological processes within living bodies and is highly applicable for clinical purposes [ 53 ]. 3D images of positron-emitting radionuclides within the body are made by a computer system. In PET-CT scanners, 3D imaging is created with the help of a CT X-ray scan implemented on the patient body in the same machine and session. Positron emission tomography (PET) and nuclear magnetic resonance (NMR) both are quantitative radiological techniques that display information about biochemistry and physiology, normality, or abnormality. Nuclear magnetic resonance is not more sensitive to give high-resolution images by the distribution of substances except hydrogen. It can measure the total concentration of ATP and creatine phosphate (CP) in particular areas of the brain. Thus, both NMR and PET performed their specific function in the diagnosis. In 1953, the first PET system was established at Massachusetts General Hospital. It was followed by many other devices in a series manner such as tomographic positron camera, PET scanner, and other PET instruments [ 54 ].

10. Working Principle of PET

The PET technique detects radioactivity emission when a small concentration of radioactive tracer is intravenously injected. These tracers are frequently labeled with carbon-11, nitrogen-13, oxygen-15, and fluorine-18, as shown in Figure 7 . There is no positron emitter of hydrogen. The radioactive dose amount is the same as used in CT. 10-40 minutes is required to complete the process to perform a PET scan. The patient is fully clothed during scanning. There are specific steps in PET scan processing, explained in Figure 7 [ 55 ].

Basic principle of PET scan: (1) a positive electron (positron) is emitted by radioactive decay of radioisotope, for example, carbon-11; (2) this positron hits an electron present in the tissue to be analyzed and emits two photons having low energy; (3) scintillation crystals are present in the PET camera to absorb this emitted photon with low energy; (4) the light is produced that is converted into another signal such as electrical signals used by the computer system to produce 3D images.

Two molecular probes are mostly used to explain PET assay: 2-[F-18] fluoro-2-deoxy-D-glucose (FDG) and 3-deoxy-3-[F-18] fluorothymidine (FLT). FDG is the analog of F-18-labeled glucose, and it is used to identify diseases by changing the metabolism of glucose in heart diseases, Alzheimer's disease, and cancer. FLT is the analog of F-18 labeled thymidine and is highly used to estimate processes like cell proliferation and DNA replication by analysis of the phosphorylation process and thymidine transport. Thus, FLT and FDG are considered as best candidate probes/tracers for molecular imaging.

An early diagnosis of Alzheimer's disease can be scanned by PET technology with 93% accuracy. Huntington's disease, a hereditary disease was also detectable by PET scan. The development of PET technology provides accurate whole-body images for examining early primary and metastatic diseases. Imaging of transgenes provides information on the regulation of gene expression during cell proliferation, growth, response to environmental stimuli, the aging process, and gene therapy. Such endogenous gene expression can be monitored through the developed PET approach, which uses F-18-labeled oligodeoxynucleotides having a short single strand of almost fifteen nucleotides. In monkeys with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine- (MPTP-) induced lesions on one side of the brain, the study of restoring dopamine production by gene therapy can be assessed by PET imaging, as shown in Figure 8 [ 56 ].

PET images of gene therapy in unilateral MPTP monkey as Parkinson's model: (a) image of normal dopamine production; (b) the image is representing autonomous dopamine MPTP-induced shortage (before gene therapy); (c) image of the restoration of dopamine production in the caudate and putamen (after gene therapy) [ 56 ].

The combination of PET and computed tomography (PET-CT) forms a hybrid imaging approach that is highly used to gain functional and metabolic information to measure inflammatory and infectious diseases to assess their proper treatment. PET-CT hybrid imaging technique can provide quick information during the diagnosis of a disease and its treatment response [ 57 ].

Haberkorn et al. measured FDG uptake that relates to the proliferation rate of tumor cells in the head and neck with different patterns, in two groups of patients [ 58 ]. The FDG uptake was also measured in the malignant neck and head tumors and metastases process by the use of FDG-PET. Minn and coworkers found that the uptake of FDG is related to the proliferation rate of tumor cells [ 59 ]. Jabour et al. measured the normal anatomy of the neck and head [ 60 ]. In this way, change in uptake FDG as investigated by PET provides necessary information for clinical and anticancer therapeutics [ 61 ].

In the human cerebellum, changes in local neuronal function by voluntary movement and tactile stimulation were also mapped with the help of the PET approach detection of brain blood flow. According to research, finger movement leads to the production of parasagittal and bilateral blood flow enhancement in the superior and anterior hemispheric cortex of the brain human brain cerebellum. The enhancement in midline blood flow in the posterior vermis of the human brain cerebellum is produced by saccadic eye movement. PET also allows the measurement of structural and functional relations in the cerebellum of the human brain [ 62 ]. The development of the PET brain imaging technology makes it possible to advance understanding of the anatomy of brain parts and map of neuroanatomical basis of cognitive processes and memory [ 63 ]. A pathogen SIV (simian immunodeficiency virus) causes infection in rhesus macaques (a type of monkeys) with acute viremia, and progression leads to infection in the solid tissues of lymphoid, and then, cellular degradation becomes a terminal disease, and death occurs in most cases. So, the FDG-PET imaging technique is used to take images from SIV-infected animals. In this way, infected groups can be distinguished from the uninfected control groups [ 64 ].

11. Future of PET Technology

The PET scan can be used to measure the concentration of amino acids, sugar, fatty acids, and receptor in the living body. It is a new diagnostic tool used to detect diseases such as atherosclerosis, aging, cancer, and schizophrenia, although improvement in instrumentation and modeling is still required for future purposes. Emission tomography has also been associated with a small risk of ionizing radiations [ 54 ].

12. Magnetic Resonance Imaging (MRI)

Magnetic resonance imaging (MRI) is primarily an imaging technique that is applicable for noninvasive visualization of the anatomy and physiology of the body in both disease and health conditions. An MRI-related technique known as echo-planar imaging (EPI) was developed by physicists Peter Mansfield and Paul Lauterbur in the late 1970s [ 65 ]. Magnetic fields, electric fields, and radio waves are used in an MRI scanner to produce images of organs and the structure of the body. The SI unit of magnetic flux density (magnetic field strength) is measured in tesla (T).

The most common detections by MRI are multiple sclerosis, CNS tumors, brain and spine infections, stroke, injuries in ligaments and tendons, muscle degradation, bone tumor, and occlusion of blood vessels. MRI uses nonionizing radiation, frequently preferable compared to CT. MRI also provides excellent contrast of soft tissues; for example, the white and gray matter structure of the brain can easily be distinguished through this approach. MRI employs other different techniques such as functional MRI, magnetic resonance angiography (MRA), susceptibility-weighted, diffusion-weighted (PWI), diffusion-weighted (DWI), gradient echo, and spin-echo. It provides an image of good quality without requiring repositioning of the patient [ 66 ]. There are several benefits to MRI such as it is a painless, noninvasive technique with high spatial resolution and nonionizing radiations. MRI is mostly used independently for soft tissue analysis.

13. Working Principle of MRI

An MRI machine consists of multiple components, including a slab for patients to lie on, a superconducting magnet, a protective cage, the operator's console, and computers to analyze the data and product images. During the MRI scanning process, the machine's magnet produces a strong magnetic field. Hydrogen ions align in the target body part of the patient due to a stable magnetic field. Then, bombardment of radiofrequency waves causes the alignment of lined-up hydrogen ions to move out, and then, ions return to their equilibrium state [ 67 ]. An attached computer system converts the spin echoes (signal) of hydrogen ions into the images, after several “shifting” and “working on.” A microphone is also present inside the MRI unit for communication between the patient and technologist during the imaging process. Images of only the target part of the body are created through MRI radiological analysis. The physician chooses which part of the patient's body must be analyzed by imaging, to diagnose the illness of the patient [ 68 ].

14. Applications of MRI

A major application of whole-body MRI is to investigate skeletal metastases. The MRI approach allows for visualization of the tumor because the tumor matrix contains an abundance of the proton. It is a more sensitive imaging technique than skeletal scintigraphy (bone scan) in the measurement of skeletal metastases. The whole-body MRI technique is more effective for detecting lesions in the pelvis, spine, and femur. This technique is also highly used as a primary diagnostic tool for the measurement of soft tissue diseases, whole-body fat, and polymyositis disease [ 69 ].

MRI is different from other diagnostic techniques because MRI has no risk of ionizing radiation. MRI has no side effects unlike CT and PET scans. There is no loss of image quality due to the scanning of body target parts from several angles and viewpoints [ 70 ]. Dynamic contrast-enhanced magnetic resonance (DCE-MRI) has been developed for the detection of the tumor microenvironment and its treatment. It has been supported as a useful method and improved clinical interest [ 71 ].

An advantage of using MRI to diagnose cardiovascular diseases is that examination reveals function structure perfusion, metabolism, and blood flow in the heart. A cardiovascular MRI is a source for the detection of congenital cardiac diseases, abnormalities in the thoracic aorta, and pericardium in heart patients. During the detection of myocardial tumor or right ventricular dysplasia, tissues are differentiated due to varying imaging parameters of the MRI approach. Another application of MRI, cardiovascular MRI, is applicable for determining cardiac prognosis, ischemia in a patient with heart disease, artery arteriosclerosis, and screening the myocardial viability [ 72 ].

Schizophrenia patients show mental abnormality which leads to language processing deficits and abnormal social behavior. Functional MRI has been applicable for remarks of such types of illnesses. The region of hypoactivity can be determined in the frontotemporal cortex of the patient brain. Soft neurological signs and symptoms have also been promoted. Functional MRI can detect abnormalities in the cerebrum; cerebral asymmetry images reveal changes in patients with schizophrenia compared with control as shown in Figure 9 [ 73 ].

Abnormal cerebral asymmetry in schizophrenia patients compared with control is shown by functional-MRI imaging technique.

Microfluidic LOC (lab-on-a-chip) is a device used as an emerging technology in medical laboratories. Sample (consist of suspensions of cells) and reagents react on these devices. For monitoring reaction on LOC, MRI is considered an ideal tool. MRI records the signals from the expended fluid leavings in the device. MRI combined with MRS (magnetic resonance spectroscopy) monitors fluid flow processes, chemical reaction separations, and diffusion processes in LOC. But MRI and MRS both show low sensitivity. In the future, there is a hope that MRI will be applicable for the advancement of microfluidic LOCs with powerful usage in medical diagnostic libraries [ 74 ].

Mutation in BRCA1 and BRCA2 genes leads to losing their ability to repair the damaged DNA, causing cancer, especially breast cancer. MRI diagnoses breast cancer which is due to a genetic mutation. Mostly, these hidden breast cancers are not detected by mammography. For a decade, doctors use MRI imaging tools to detect breast cancer [ 75 ]. According to previous research, 27-37% of patients have shown lesions on MRI, which are not seen through mammography. The researchers noted that mammography had a low value of positive prediction of 52.8%, as compared to MRI which is high at 72.4% [ 74 ].

Molecular MRI employed for specific and early detection of pulmonary metastatic cancer cells can improve its treatment. In research, pulmonary cancer cells are besieged by iron oxide nanoparticles having the ability to bind with ligand expressed on the cells. Then, images were taken by high-resolution hyperpolarized 3 He MRI (HP 3 He MRI). The study confirmed that HP 3 He MRI pooled with targeted superparamagnetic iron oxide nanoparticle (SPION) contrast agent detects specific and early metastatic pulmonary cancer in mice. A researcher used the LHRH-SPION agent to explore new drug procedures. For this purpose, they injected breast adenocarcinoma cells into mice and then detect pulmonary defects as cancer formation in mouse lungs with the help of LHRH-SPIONs and HP 3 He MRI results are shown in Figure 10 [ 76 ].

Breast adenocarcinoma mouse model formed for the detection of lung metastases. (a) High-resolution hyperpolarized 3 He MRI (HP 3 He MRI) images were taken from the control mouse. Screening as normal ventilation forms of lungs. (b) After injection of LHRH-SPIONs, images were produced from human breast adenocarcinoma abnormal mouse (model), showing defects in the right lobe (under circles).

Multinuclear 3D solid-state MRI allows images of tooth bone and calcium phosphate components of bone substances. It also gives information on the bone composition and texture of the bone [ 77 ]. Recent neuroimaging techniques including high-resolution MRI can investigate myeloarchitectural patterns in the cortex of the human brain. The bands of myelination have been revealed by the staining technique. Now, the same band in good quality image form can also be obtained by high-resolution MRI imaging technique. Although the advanced technology has been largely applied in the visual system, further improved methodologies are required for the investigation of another brain region. To overcome high ratio of “signal” to “noise” is a challenge for MRI machines that is produced due to the increased resolution of the image [ 78 ]. In this way, fMRI and other types of MRI have been powerfully used to change our understanding of diseases, their causes, and how to manage the conditions.

15. Simultaneous Imaging with MRI and PET

In in vivo study, imaging of small animals, for example, mouse imaging by combined MRI and PET modalities, produces constant information of different parts of the body. An experimental study reveals that the combined PET/MRI technique improves the understanding of malignant tissues and heterogeneous tumors, edema, and necrosis that are not done by MRI alone. Particular ionic 18 F is also used for PET/MRI combined imaging of small animals as models of bone metastasis, osteoporosis, and arthritis to study the complete skeletal system [ 79 ].

16. Risks of Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) is highly expensive, low in sensitivity, and time-consuming for scanning and processing compared to other imaging modalities. A probe with a bulk quantity may be needed for MRI. It cannot detect abnormalities of intraluminal body parts. It gives no real-time information. And it can create a suffocating environment for some people [ 2 ].

17. Single-Photon Emission Computed Tomography (SPECT)

Single-photon emission computed tomography (SPECT) is an advanced imaging technique using gamma rays and provides three-dimensional (3D) representations of objects with high accuracy. In 1963, Kuhl and Edwards [ 80 ] gave the first report about single positron emission computed tomography. Gradually, modification with new instruments such as computer-attached systems and rotating gamma cameras leads to the development of a novel modality of single-photon emission tomography.

SPECT has become a great medical imaging technique used in research and clinical area. A dual-headed single-photon emission tomography (SPECT) system has been shown in Figure 11 [ 81 ]. It monitors the 3D information of an object by producing series of thin slices from tomographic images. These essential tomographic images can improve the ability detection of deep and very small fractures in patients [ 81 ].

Dual-headed single-photon emission computed tomography (SPECT) system.

SPECT assesses the multiple two-dimensional (2D) images from different angles by using high-energy gamma rays. Data is reconstructed and recorded, and 3D images of the target portion of the body are produced by a computer program. SPECT is greatly used in clinics and research laboratories like other tomographic modalities such as PET, MRI, and CT. SPECT and PET both use the radioactive tracer and then measure the emitted gamma rays. In the case of SPECT, emitted gamma rays by radioactive tracers are directly detected by the detector. The computer system analyzes the data from the detector and produces the true image of the area where the radioactive tracers are injected. SPECT imaging technique is less expensive than exclusively used for imaging small animals. It is sensitive to monitoring target bone metabolism, myocardial disorders, and blood flow in the cerebrum [ 82 ].

SPECT has also been designed for imaging of the brain known as neurochemical brain imaging. It has a powerful imaging technique to elaborate the neuropsychiatric diseases. It is an essential developmental technique that has great potential to monitor the pathophysiology and many other complex disorders of the brain [ 83 ].

Today, hybrid SPECT and CT are progressively employed and available in the nuclear medicine field. SPECT/CT provides exact abnormal bone turnover during inflammation, bone tumor, bone regeneration, bone infection, and trauma in complex bone joints such as the knee, hip, foot, shoulder, and hand/wrist. In most cases, CT with specificity and SPECT with high sensitivity are performed together for a complete diagnosis. SPECT/CT is also responsible for giving the proper information about therapy planning. For example, it tells us about the decision of using SPECT/CT alone or joint arthroplasty. It was observed that joint arthrography of the knee gives better results as compared to SPECT/CT alone as shown in Figure 12 [ 84 ].

(a) The SPECT/CT scan of the knee joint shows clear articular cartilage. (b) The SPECT/CT arthrography image of the knee joint showing enhance the value of SPECT/CT screening.

SPECT/CT technique has great potential for the measurement of articular cartilage, soft bodies, and synovial structures. It also gives promising results in the examination of osteochondral abnormalities [ 84 ].

18. Risks of Single-Photon Emission Computed Tomography (SPECT)

Single-photon emission computed tomography (SPECT) is expensive and requires additional care for performance with radioactive materials. Like PET, it uses ionizing radiation and creates radiation side effect for the patients [ 2 ].

19. Digital Mammography

Mammography is a technique used to screen and diagnose human breasts. Low-energy X-rays (30 kVp) are used in the mammography in performing early diagnosis and screening of human breast cancer. Ionizing radiations are used in mammograms to create images for analysis of abnormal conditions. Ultrasound is usually employed to give additional information about masses, detected by mammography. MRI, discography, and positron emission mammography (PEM) are also supporters of mammography. Mammography is more accurate for women 50 or above 50 years old as compared to younger women because old women have high breast density [ 85 ]. Today, conventional mammography has been replaced by digital mammography.

An advanced technique is used for creating 3D images of breast tissues for detailed analysis of breast cancer, known as 3D mammography. When 3D mammography is used along with usual mammography, it gives more positive results [ 86 ]. Cost-effectiveness and high radiation exposure are of high concern to 3D mammography [ 87 ].

Digital mammography is a special form of mammography employed to investigate breast tissues for breast tumor study. Digital mammography contrasts with film (conventional) mammography by using a special detector that detects the transmitted X-rays energy and converts it into an image signal by a computer system instead of a film X-ray. Digital mammography is a rapid and advanced modality that has the potential for diagnosis and proper screening of breast cancer. Such new diffusion technologies can alter the health care pattern through many mechanisms. There are different results related to breast health care for digital-screen and film mammography. Digital mammography may also change the application of diagnostic services ensuring mammograms with positive screening [ 88 ].

Digital mammography has been considered a better technique as compared to film mammography in the detection of breast cancer in premenopausal, premenopausal, and young women. A digital system has more cost (approximately 1.5 to 4 times) than a film system. Digital mammography has the advantage of diagnosis in computer-based system that generates images with easy access and better-quality transmission, recovery, and image storage. Advanced digital mammography uses an average low dose of radiation without cooperation with diagnostic accuracy [ 89 ].

A healthy breast tissue mammogram recorded by digital mammography is clear as compared to a mammogram on film as shown in Figure 13 [ 90 ]. Scientists used both digital mammography and film mammography for 42,760 women's breast X-rays. Cancer is almost equally well detected through these techniques. But digital mammography detected 28% more breast cancer in younger women or those under 50, who have dense breast tissues. Digital mammography uses a specific detector that captures the transmitted X-rays and sends the information in the form of energy that converts into an image through a computer system [ 90 ]. Digital mammography screening for breast cancer is not cost-effective, relative to conventional mammography [ 91 ].

(a) A healthy breast tissue mammogram recorded by digital mammogram; (b) the same breast tissue mammogram recorded by film mammography. White spots shown in the above images are deposits of calcium which can consider the mark of cancer when they form clusters.

Now, screening MRI is highly used as an adjunct to mammography, recommended for women to ensure 20-25% or more threat of breast cancer [ 92 ]. Currently, breast cancer diagnostic programs have been recognized widely at least in 22 countries. Collective struggles lead to the formation of the Breast Cancer Screening Network (IBSN), an international program for policymaking, administration, funding, and handling of results that come from breast cancer screening of huge populations [ 93 ]. The ultrasound technology combined with mammography is used to detect elevated risk factors for breast cancer. According to research on elevated risk factors, by using ultrasound with mammography, more than 90% of cancer risk was seen in over 50% of women having dense breast tissues, and 25% of cancer risk factors were determined in women having just 26%-40% dense tissues of the breast. It is suggested that screening with ultrasound may be beneficial for those women having other risk factors and less dense breast tissues [ 94 ].

20. Medical Ultrasound

The ultrasound imaging technology was used earlier as a diagnostic tool for brain images. Today, ultrasound is a widespread imaging technology used in diagnostic laboratories and clinics. It is free from radiation exposure risk, comparatively less expensive, and highly portable as compared to other imaging techniques like MRI and CT [ 95 ]. This system is used in different fields. In the medical field, ultrasound uses sound waves of high frequency, to diagnose the organs and structure of the body. The ultrasound system is performed with high frequency. Special technicians or doctors use it to observe the kidney, heart, liver, blood vessels, and other organs of the body. The most critical component of ultrasound is a transducer. An ultrasound transducer can convert an electrical signal into sound waves and sound waves into an electrical signal. An ultrasonic image or sonogram is formed by transmitting pulses of sound waves into tissues using a special probe. Different tissues reflect these sounds to a different degree. These reflected sound waves (echoes) are detected and presented as an image with the help of the operator. In the medical field, there are many applications of ultrasound. Ultrasound scanning is a very effective and reliable technique that is greatly used for monitoring normal pregnancy, placenta previa, multiple pregnancies, and different abnormalities during pregnancy and rest [ 96 ].

An ultrasound imaging technique also known as transvaginal ultrasound (TVS) alters our understanding of the management and diagnosis of pregnancy. The TVS gives clear knowledge of early pregnancy problems. It investigates the pregnancy location as well as viability. In utero , the TVs have determined the fetal heart activity that is initial proof of pregnancy viability. Abnormal development in fetal heart rate pattern indicates the subsequent miscarriage. Less fetal heart rate especially at 6 to 8 weeks demonstrates subsequent fetal disease. Fetal heart pulsation can be seen on TVS. The routine use of TVS also leads to development in managing early pregnancy failure. Awareness of pregnant women and improvement in early pregnancy units can directly manage miscarriage. TVS is a very sensitive approach for diagnosing early miscarriage. It detects trophoblastic tissue and blood flow in the intervillous space, and the use of color Doppler images leads to estimate the level of expectant management [ 97 ].

Functional ultrasound (fUS) is highly applicable for imaging the brain and detects transient alternation of blood volume in the brain at high resolution than other brain imaging modalities. The blood volume in small vessels can be measured by functional ultrasound (fUS), which uses plane-wave illumination with a high frame rate. Functional ultrasound can detect the brain's active portion [ 98 ]. Ultrasound has major advantages as noninvasive and out-patient scanning in children to investigate neuromuscular disorders. Ultrasonography has been used for finding the normal function of muscle, muscle contraction, muscle thickness, and muscle fiber length. Real-time ultrasound is used for muscle imaging. When ultrasound applies to neuromuscular patients, different muscle disorders can be detected by a pattern of muscle echo, such as a bright spotted pattern of increase in muscle echo obtained in muscular dystrophy patients and a moderate increase in echo showed in spinal muscular atrophy [ 99 ].

Currently, early care physicians can easily understand complex patient conditions with the help of advanced 3D ultrasound algorithms. And high-speed networks are used in special health centers to enhance patient care facilities [ 100 ]. 3D ultrasound is widely used due to the reason of 2D ultrasound limitations. Clinical 3D ultrasound experience has an advantage in the diagnosis of disorders and produces a 3D image that guides invasive therapy. Further improvement in 3D imaging software, as well as hardware, will lead to routine usage of this tool [ 101 ].

High-resolution ultrasound which is also known as ultrasound biomicroscopy (UBM) has clinical applications in imaging the human eye. UBM uses 35 MHz or above frequencies to provide images of high resolution as compared to conventional ophthalmic ultrasound techniques. UBM can be used to diagnose ocular trauma and complex hypotony. It can determine eye lens displacement, iridodialysis, zonular flaw, cataract, lens subluxation, and hyphemia. Hyphemia is a condition in which blood diffuses the anterior chamber of the eye due to injury. Hyphemia scan image by Sonomed UBM is shown in Figure 14 [ 102 ].

(a) Hyphemia is due to injury; blood diffuses the anterior chamber of the eye. (b) Hyphemia scan image by Sonomed UBM.

UBM can detect several eye abnormalities, as it allows for the diagnosis of trauma, glaucoma, and foreign bodies. It can evaluate eyelids, neoplasms, normal eye anatomy, and extraocular muscles during strabismus surgery. Currently, for high-resolution diagnostic eye imaging, new advanced technologies such as pulse encoding, transducer array, and ultrasound combination with light are developed [ 102 ].

A high-frequency (40 MHz) ultrasound imaging technique has been developed for checking the programmed cell death process known as apoptosis. It is a noninvasive procedure used to monitor the apoptosis process that happens because of agents especially anticancer agents in cells of body tissue in vitro or in vivo. The procedure of detection monitored alternations in subcellular nuclear such as the condensation process after proper destruction of the cell during the programmed cell death process. The high-frequency ultrasound technique shows a high scattering rate due to these intense alternations (approximately 25-50-fold) in apoptotic cells as compared to normal tissue cells. As a result, the apoptotic tissues show greatly brighter areas as compared to normal tissue. In the future, this noninvasive imaging technique will use to check the effect of anticancer treatment and chemotherapeutic agents in laboratory model systems and then in patients [ 103 ].

Another application of ultrasound technology is the successful detection of renal masses. According to research results, 86% carcinomas and 98% renal cysts were accurately determined among 111 patients by the ultrasound imaging technique. Ultrasound is a safe, simple, and cheap diagnostic tool to diagnose complex renal masses [ 104 ]. Ultrasound screening allows imaging the cartilage for checking instability, abnormal location of the femoral head inside the acetabulum, and developmental dysplasia in newborns [ 105 ]. A powerful low-frequency ultrasound system can also be used as a noninvasively drug delivery system. Many drugs and proteins having high molecular weight can be delivered easily with excessive permeability into human skin with the help of a low-frequency ultrasound modality. For example, insulin, erythropoietin, and interferon-gamma molecules are easily and safely delivered across human skin [ 106 ].

A novel molecular imaging technique known as molecular ultrasound has been used by researchers in the molecular biology field to monitor the alternation in the expression rate of molecular markers located on intravascular targets. Contrast agents used in the advanced molecular ultrasound imaging technology are mostly micro- or nanosized particles having ligands on the surface also known as microbubbles. Specific molecular markers are targeted by these microbubbles such as selectin, vascular cell adhesion molecule 1, and integrin. In this way, these agents lead to detect specific molecular markers on intravenous targets by gathering at that specific tissue site.

Molecular ultrasound has many advantages such as low-effective cost, high resolution, portable, noninvasiveness, absence of ionizing radiation, real-time imaging potential, and high availability. It has the potential for regular investigations of different abnormalities at the molecular level, such as inflammation, tumor angiogenesis, and thrombus. In addition, improvement is still required in the field of molecular ultrasound to design the novel targeting ligand to form a more effective contrast agent. In the future, advancements in molecular ultrasound imaging technology will play a clinical role with high sensitivity and accuracy for imaging complex abnormalities at the molecular level [ 107 , 108 ].

21. Disadvantages of Medical Ultrasound

There are many useful applications of ultrasound in the medical field. But medical ultrasonography also had some side effects such as hormone change effect, breakage of chromosomes with very low frequency, chemical effects, and other health problems. It was examined that for 10-week gestation, chorionic gonadotropin hormone in humans increased after routine usage of ultrasound modality [ 109 ].

It was observed in an experiment that no chromosome damage occurred after diagnostic ultrasound exposure to human lymphocyte culture. But experimental results suggest that ultrasound can cause chromosome breakage with very low frequency [ 110 ]. Similarly, lots of experiments have been done to check fetal ultrasound safety. In vivo study demonstrates no major neurological defects, fetal growth problems, birth flaws, or childhood cancer caused by ultrasound imaging. But in vitro study demonstrates the possibility of some health problems that can occur due to diagnostic ultrasonography. For example, the effect of diagnostic ultrasound on the normal architecture of the mouse fibroblast cell with the production of fingerlike projections on the fine surface of the cell is demonstrated in Figure 15 [ 111 ].

Images of mouse fibroblast cell (a) normal smooth form and (b) abnormal cell in its rough shape. Due to the diagnostic ultrasound effect, fingerlike projections are produced on the smooth surface of the fibroblast cell.

Another disadvantage of an ultrasound system in the medical field is generating significant heat. When strong ultrasound waves migrate through the liquid environment, small cavities are produced, and these cavities expand, then collapse to each other, and generate heat. This heat-generating condition leads to create an unnecessary chemical environment [ 112 ]. Dyslexia, growth limitation, non-right handedness, and late speech-like effects are also examined after diagnostic ultrasound exposure. Continuous research is required to find the side effects of medical ultrasonography on human health [ 113 ].

22. Radiation Exposure Risk from Medical Imaging and Its Management

Radiological imaging by X-ray radiology, positron emission tomography (PET), single-photon emission computed tomography (SPECT), mammography, and computed tomography (CT) uses high-energy X-rays that leads to high radiation dose in some patients. Pregnant women and children are mostly affected by radiation exposure from radiological imaging. Possibility of cancer, genetic mutations, growth and developmental retardation in the fetus, and cardiovascular abnormalities can occur by the exposure to radiation after radiotherapy. Direct radiation exposure can cause hair loss, cataracts, skin redness, and skin damage. Radiation exposure risks can be reduced by making “National Guidelines” that aid the physician to manage their effects on patients.

Many online tools have been developed to enable physicians or technicians to record the calculation of radiation exposure from each radiological imaging technique. For this purpose, a Thermoluminescent dosimeter (TLD) is used to calculate radiation dose through different software depending on the modality being used such as Monte Carlo PENRADIO which is used for CT [ 114 ]. Magnetic resonance imaging (MRI) and ultrasound techniques are free from ionizing radiation. To minimize the risk, MRI and ultrasound can be used instead of radiological imaging techniques. Reduction in unnecessary computed tomography screening leads to the direct reduction of radiation exposure risks. Today, advanced and safe technologies are used that allow measuring signals with a low dose of radiation; e.g., low-dose computed tomography scanners permit less radiation exposure [ 115 ].

23. Advanced Machine Medical Image Analysis

4D medical imaging analysis is an advanced technology that is used in combination with different modalities such as 4D CT, 4D US, and 4D MRI. 4D CT is an excellent choice for radiation oncology, which is prone to motion artifacts. Similarly, 4D ultrasound is particularly useful in prenatal research. 4D flow MRI can help doctors diagnose and treat heart issues more precisely. For big data integration in medical imaging, researchers need to develop algorithms to store images by converting them into numerical format which will be helpful for physicians in diagnosis.

24. Artificial Intelligence in Medical Imaging

Machine learning and deep learning are branches of artificial intelligence (AI) that solves problems in medical imaging applications such computer-aided diagnosis, lesion segmentation, medical image analysis, image-guided treatment, annotation, and retrieval. AI assesses image quality, interprets image, and analyzes biomarkers, and finally, reporting is done. AI has impact on oncologic imaging. Lung cancer is one of the most prevalent and severe tumors in thoracic imaging. AI can assist in recognizing and classifying these nodules as benign or cancerous [ 116 ]. Machine learning focuses on pattern recognition. The traditional AI systems relied on preset engineering feature algorithms with specified parameters based on expert knowledge. Such features were intended to assess certain radiographic characteristics such as a tumor's 3D shape and intratumoral texture. Following that, a selection process ensures that only the most important features are used. The data is then fed into statistical machine learning models to find potential imaging-based biomarkers [ 117 ]. Deep learning algorithms learn by navigating the data space and by providing them with greater problem-solving capabilities. Convolutional neural networks (CNNs) are the most used deep learning architectural typologies in medical imaging today, even though many deep learning designs have been researched to handle diverse objectives [ 118 ].

25. Critical Analysis

Medical imaging often contains several techniques that are noninvasive to make images of different parts of the body. New imaging techniques such as computed tomography (CT), positron emission computed tomography (PET), magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT), ultrasound (US), and digital mammography reveal the internal anatomy and physiology of the body [ 2 ].

Advanced imaging technologies are used to diagnose various external as well as internal human illnesses that can also minimize diagnostic errors and produce novel and better information about the target object. There are benefits and risks to every imaging technique. Ultrasound is also employed in the medical field to look at the kidneys, heart, liver, blood vessels, and other organs of the body [ 95 ]. Computed tomography (CT) measures cancer and various abnormalities in the heart, abdomen, bone, and spinal cord with high resolution [ 33 ]. 3D ultrasound computed tomography (3D USCT) is a promising technology for the investigation of breast cancer [ 48 ]. PET is a powerful technique to visualize, characterize, and quantify the biological processes and pathological changes at the cellular and subcellular levels within a living body. The development of MRI is now employed to examine several musculoskeletal, neurologic problems, and cancer. It can be used for both soft and hard tissues [ 1 ]. Digital mammography is a rapid and computer-based modality. It is used for the diagnosis and screening of breast cancer [ 89 ].

Some imaging techniques, such as CT, PET, SPECT, and digital mammography using X-rays, lead to high ionizing radiation exposure risk in some patients. There are some management steps for minimizing the radiation exposure risks from imaging techniques [ 114 ]. The development in medical imaging techniques such as the use of PET/CT hybrid, SPECT/CT hybrid, 3D USCT, and simultaneous PET/MRI leads to an increase in our understanding of diseases and their treatment [ 84 ]. With advanced medical imaging techniques, detection of early-stage diseases is possible and then eventually aids patients to live longer and better lives. In the future, with mounting innovations and advancements in technology systems, the medical diagnostic field would become a field of regular measurement of various complex diseases and will provide healthcare solutions [ 1 ].

Acknowledgments

The authors wish to thank Research Center College of Pharmacy at King Saud University, Riyadh, Saudi Arabia, for their financial support and for providing free access to digital library and laboratory.

Abbreviations

Data Availability

Conflicts of interest.

The authors declare no conflicts of interest.

Authors' Contributions

All authors contributed equally.

Characteristic curve

• Report problem with article
• View revision history

Citation, DOI, disclosures and article data

At the time the article was created Michael Francois Nel had no financial relationships to ineligible companies to disclose.

At the time the article was last revised Raymond Chieng had no financial relationships to ineligible companies to disclose.

• Characteristic curves
• H and D curve
• Hurter and Driffield curve
• H and D curves

The characteristic curve , also known as the H and D curve , is a representation of the response of a screen film radiograph to light. The characteristic curve represents the change in optical density (OD) of the screen film in response to changing exposures (incident x-rays on the screen film).

On this page:

Image formation, optical density, relative exposure.

• Cases and figures

Incident x-rays reaching the screen film are converted to light by the scintillator (commonly Gd 2 O 2 S crystals) within the intensifying screen , a process called fluorescence . X-ray conversion to light is directly proportional.

Light from the intensifying screen interacts with the silver halide molecules in the emulsion that coats the screen film base. The silver (Ag + ) ions undergo reduction when exposed to light. This gain of an electron converts Ag + into a stable uncharged Ag atom. The deposition of silver atoms onto the film represents the latent image.

The film is then processed . The stable silver atoms act as a catalyst for surrounding Ag ions to be reduced. Silver deposition onto the film represents dark regions on the film. More silver deposition on the film results in less light passing through that region when placed on a light box .

ADVERTISEMENT: Supporters see fewer/no ads

The amount of light permitted through the film is known as the transmittance value . Transmittance (T) is calculated as follows 4 :

T = I t /I 0

I t = intensity of light transmitted through the screen film

I 0 = Intensity of light from the light box

Meanwhile, I 0 /I t measures the opacity of the film or ability of the film to stop light 4 . The transmittance value can be converted to a unit known as optical density (OD) 4 .

OD = -log 10 (T) or log 10 (I 0 /I t )

A change in OD of 1 represents a 10-fold change in the transmittance of light.

OD never reaches 0. A new film will have a minium OD of base + fog of 0.12 4 .

At the "toe" and "shoulder" regions of the characteristic curve, a large change in relative exposure does not produce a significant change in OD 4 .

Exposure of a radiographic film is measured in mAs and can be adjusted by varying the time of exposure. Relative exposure is the amount of exposure increase needed to produce a change in optical density. Doubling the amount of exposure will double the change in optical density. The absolute exposure (measured in mAs or mR or number of photons per square mm) is not needed to understand the relationships between exposures 4 . Using logarithmic scale of relative exposure has several adavantages, namely: wide range of exposures can be represented on a single graph and exposures can still be represented on a linear scale, making curve analysis easier 4 .

An increase in log of exposure of 0.3 represents the doubling of relative exposure 4 .

Latitude is the range of logarithmic exposures that produces useful OD (usually OD of between 0.25 to 2.0) 4 .

The characteristic curve is specific for each screen film. The linear part of the characteristic curve represents the exposure range at which OD will change linearly. This exposure range is equivalent to the dynamic range/latitude of the screen film 5,6 . The steeper the linear region, the less dynamic range the film has, and the more contrast the radiograph will have 4 .

The slope or gradient of the linear portion of the curve is known as film gamma, given by the formula below 4 :

Gamma = (D 2 -D 1 ) / (Log E 2 - Log E 1 )

where D 2 -D 1 is the difference between the optical densities of the steepest part of the curve while Log E 2 - Log E 1 is the difference between the exposures 4 . A gamma of one represents the original subject contrast. Meanwhile a gamma of more than one exaggerates the subject's contrast and a gamma of less than one decreases the subject's contrast 4 .

A film-screen system will always have higher contrast than films that exposed to X-rays directly. This is because an intensifying screen always have higher sensitivity towards X-rays. An intensifying screen would require 1 mR of exposure to produce a density of 1.0 while exposing films directly to X-rays would require 30 mR of exposure 4 .

Film speed or sensitivity is the exposure required (in roentgens) to produce a density of 1.0 above base + fog density. It can be defined as follows 4 :

Speed = 1/roentgens

The difference between speed and gamma is that: gamma affects the film contrast and the shape of the characteristic curve. Meanwhile, the speed affects the location of the curve along the logarithmic exposure scale 4 :

A film that requires higher exposures to change OD values (a right shift of the characteristic curve) is said to be a 'slower' or 'less sensitive' film 3 . Increasing the development time or temperature increases the film speed/sensitivity, film gamma and fogging because less exposure is needed to produce a specific change in OD values 3,4 .

Higher speed film tends to produce more noise 2 .

Quiz questions

• 1. Jerrold T. Bushberg, J. Anthony Seibert, Edwin M. Leidholdt et al. The Essential Physics of Medical Imaging. (2011) ISBN: 9781451153941 - Google Books
• 2. LePage J & Trefler M. The Use of High-Speed Film/Screen Combinations to Improve Diagnostic Image Quality. Radiology. 1983;147(1):265-7. doi:10.1148/radiology.147.1.6828744 - Pubmed
• 3. Perry Sprawls. Physical Principles of Medical Imaging. (1993) ISBN: 9780834203099 - Google Books The Photographic Process and Film Sensitivity
• 4. Thomas S. Curry, James E. Dowdey, Robert C. Murry. Christensen's Physics of Diagnostic Radiology. (1990) ISBN: 9780812113105 - Google Books
• 5. Williams M, Krupinski E, Strauss K et al. Digital Radiography Image Quality: Image Acquisition. Journal of the American College of Radiology. 2007;4(6):371-88. doi:10.1016/j.jacr.2007.02.002 - Pubmed
• 6. General screen-film radiography and its limitations. University of Sydney. Chapter 2

Incoming Links

• Characteristic curves of high and low speed films
• Question 2746
• Question 2745

Promoted articles (advertising)

By section:.

• Artificial Intelligence
• Classifications
• Imaging Technology
• Interventional Radiology
• Radiography
• Central Nervous System
• Gastrointestinal
• Gynaecology
• Haematology
• Head & Neck
• Hepatobiliary
• Interventional
• Musculoskeletal
• Paediatrics
• Not Applicable

Radiopaedia.org

• Feature Sponsor
• Expert advisers

• Radiology Key

Fastest Radiology Insight Engine

• BREAST IMAGING
• CARDIOVASCULAR IMAGING
• COMPUTERIZED TOMOGRAPHY
• EMERGENCY RADIOLOGY
• FETAL MEDICINE
• FRCR READING LIST
• GASTROINTESTINAL IMAGING
• GENERAL RADIOLOGY
• GENITOURINARY IMAGING
• HEAD & NECK IMAGING
• INTERVENTIONAL RADIOLOGY
• MAGNETIC RESONANCE IMAGING
• MUSCULOSKELETAL IMAGING
• NEUROLOGICAL IMAGING
• NUCLEAR MEDICINE
• OBSTETRICS & GYNAECOLOGY IMAGING
• PEDIATRIC IMAGING
• RADIOGRAPHIC ANATOMY
• RESPIRATORY IMAGING
• ULTRASONOGRAPHY
• Abdominal Key
• Anesthesia Key
• Basicmedical Key
• Otolaryngology & Ophthalmology
• Musculoskeletal Key
• Obstetric, Gynecology and Pediatric
• Oncology & Hematology
• Plastic Surgery & Dermatology
• Clinical Dentistry
• Thoracic Key
• Veterinary Medicine
• Gold Member
• iOS/Android App

Digital Radiographic Image Processing and Manipulation

Chapter 3 Digital Radiographic Image Processing and Manipulation Outline PSP Reader Functions Image Sampling Digital Radiography Image Sampling The Nyquist Theorem Aliasing Automatic Rescaling Look-Up Table Quality Control Workstation Functions Image Processing Parameters Contrast Manipulation Spatial Frequency Resolution Spatial Frequency Filtering Basic Functions of the Processing System Image Manipulation Image Management Patient Demographic Input Manual Send Archive Query Summary Objectives On completion of this chapter, you should be able to: •  Describe the formation of an image histogram. •  Discuss automatic rescaling. •  Compare image latitude in digital imaging with film/screen radiography. •  List the functions of contrast enhancement parameters. •  State the Nyquist theorem. •  Describe the effects of improper algorithm application. •  Discuss the purpose and function of image manipulation factors. •  Describe the major factors in image management. Key Terms Aliasing Archive query Automatic rescaling Contrast manipulation Critical frequency Edge enhancement High-pass filtering Histogram Image annotation Image orientation Image sampling Image stitching Look-up table (LUT) Low-pass filtering Manual send Nyquist theorem Patient demographics Shuttering Smoothing Spatial frequency resolution Window level Window width Once x-ray photons have been converted into electrical signals, these signals are available for processing and manipulation. This is true for both photostimulable phosphor (PSP) systems and flat-panel detector (FPD) systems, although a reader is used only for PSP systems. Processing parameters and image manipulation controls are also similar for both systems. Preprocessing takes place in the computer where the algorithms determine the image histogram. Postprocessing is done by the technologist through various user functions. Digital preprocessing methods are vendor specific, so only general information on this topic can be covered here. PSP Reader Functions The PSP imaging plate records a wide range of x-ray exposures. If the entire range of exposure were digitized, values at the extremely high and low ends of the exposure range would also be digitized, resulting in low-density resolution. To avoid this, exposure data recognition processes only the optimal density exposure range. The data recognition program searches for anatomy recorded on the imaging plate by finding the collimation edges and then eliminates scatter outside the collimation. Failure of the system to find the collimation edges can result in incorrect data collection, and images may be too bright or too dark. It is equally important to center the anatomy to the center of the imaging plate. This also ensures that the appropriate recorded densities will be located. Failure to center the imaging plate may also result in an image that is too bright or too dark. The data within the collimated area produce a graphic representation of the optimal densities called a histogram ( Figure 3-1 ). The value of each tone is represented (horizontal axis), as is the number of pixels in each tone (vertical axis). Values at the left represent black areas. As tones vary toward the right, they get brighter, with the middle area representing medium tones. The extreme right area represents white. A dark image will show the majority of its data points on the left, and a light image will show the majority of its data points on the right. FIGURE 3-1 Simple Histogram Illustration. A graphical representation of the number of pixels with a particular intensity. Because the information within the collimated area is the signal that will be used for image data, this information is the source of the vendor-specific exposure data indicator. Image Sampling With image sampling , the plate is scanned and the image’s location is determined. The size of the signal is then determined, and a value is placed on each pixel. A histogram is generated from the image data, which allows the system to find the useful signal by locating the minimum (S1) and maximum (S2) signal within the anatomic regions of interest on the image and then plots the intensities of the signal on a histogram. The histogram identifies all intensities on the imaging plate in the form of a graph on which the x -axis is the amount of exposure read, and the y -axis is the number of pixels for each exposure. This graphic representation appears as a pattern of peaks and valleys that varies for each body part. Low energy (low kilovoltage peak [kVp]) gives a wider histogram; high energy (high kVp) gives a narrower histogram. The histogram shows the distribution of pixel values for any given exposure. For example, if pixels have a value of 1, 2, 3, and 4 for a specific exposure, then the histogram shows the frequency (how often they occurred) of each of those values, as well as the actual number of values (how many were recorded). Analysis of the histogram is very complex. However, it is important to know that the shape of the histogram is anatomy specific, which is to say that it stays fairly constant for each part exposed. For example, the shape of a histogram generated from a chest x-ray on an adult patient will look very different than a knee histogram generated from a pediatric knee examination. This is why it is important to choose the correct anatomic region on the menu before processing the image plate. The raw data used to form the histogram are compared with a “normal” histogram of the same body part by the computer, and the image correction takes place at this time ( Figure 3-2 ). FIGURE 3-2 The histogram shows the pixel values found within this chest x-ray. (Courtesy American Association of Physicists in Medicine.) Digital Radiography Image Sampling The Nyquist Theorem In 1928 Harry Nyquist, who was a researcher for AT&T, published the paper “Certain Topics in Telegraph Transmission Theory.” He described a way to convert analog signals into digital signals that would more accurately transmit over telephone lines. He found that since an analog signal was limited to specific high frequencies, it could be captured and transmitted digitally and recreated in analog form on the receiver. He said that the sampling rate would need to be at least twice the highest frequency to be reproduced. In 1948 Claude Shannon presented a mathematical proof of Nyquist’s theory, allowing it to be called the Nyquist theorem. Since that time, a number of scientists have added to and revised the theory. In fact, it could be called the Nyquist–Shannon–Kotelnikov, Whittaker–Shannon–Kotelnikov, Whittaker–Nyquist–Kotelnikov–Shannon (WNKS), etc., sampling theorem, as well as the Cardinal Theorem of Interpolation Theory. It is often referred to simply as the sampling theorem. The Nyquist theorem states that when sampling a signal (such as the conversion from an analog image to a digital image), the sampling frequency must be greater than twice the frequency of the input signal so that the reconstruction of the original image will be as close to the original signal as possible. In digital imaging, at least twice the number of pixels needed to form the image must be sampled. If too few pixels are sampled, the result will be a lack of resolution. At the same time, there is a point at which oversampling does not result in additional useful information. Once the human eye can no longer perceive an improvement in resolution, there is no need for additional sampling. The number of conversions that occur in PSP imaging—electrons to light, light to digital information, digital to analog signal—results in loss of detail. Light photons do not travel in one direction, so some light will be lost during the light-to-digital conversion because light photons spread out. Because there is a small distance between the phosphor plate surface and the photosensitive diode of the photomultiplier, some light will spread out there as well, resulting in loss of information. In addition, even though the imaging plate is able to store electrons for an extended period of time, the longer the electrons are stored, the more energy they lose. When the laser stimulates these electrons, some of the lower energy electrons will escape the active layer, but if enough energy was lost, some lower energy electrons will not be stimulated enough to escape and information will be lost. All manufacturers suggest that imaging plates be read as soon as possible to avoid this loss. Although FPD systems lose fewer signals to light spread than PSP systems, the Nyquist theorem is still applied to ensure that sufficient signal is sampled. Because the sample is preprocessed by the computer immediately, signal loss is minimized but still occurs. Aliasing

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Stay updated, free articles. Join our Telegram channel

Comments are closed for this page.

Full access? Get Clinical Tree

Fundamental Properties of X-rays

• First Online: 01 January 2011

Cite this chapter

• Yoshio Waseda 4 ,
• Eiichiro Matsubara 5 &
• Kozo Shinoda 4  

13k Accesses

2 Citations

X-rays with energies ranging from about 100eV to 10MeV are classified as electromagnetic waves, which are only different from the radio waves, light, and gamma rays in wavelength and energy. X-rays show wave nature with wavelength ranging from about 10 to 10 −3 nm. According to the quantum theory, the electromagnetic wave can be treated as particles called photons or light quanta. The essential characteristics of photons such as energy, momentum, etc., are summarized as follows.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or eBook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime
• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Available as EPUB and PDF
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Author information

Authors and affiliations.

Institute of Multidisciplinary Research for Advanced Materials, Tohoku Uninversity, Katahira 2-1-1, Aoba-ku, 980-8577, Sendai, Japan

Professor Dr. Yoshio Waseda & Professor Kozo Shinoda

Graduate School of Engineering Department of Materials Science and Engineering, Kyoto University, Yoshida Honmachi, Sakyo-ku, 606-8501, Kyoto, Japan

Professor Dr. Eiichiro Matsubara

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Yoshio Waseda .

Rights and permissions

Reprints and permissions

Copyright information

© 2011 Springer-Verlag Berlin Heidelberg

About this chapter

Waseda, Y., Matsubara, E., Shinoda, K. (2011). Fundamental Properties of X-rays. In: X-Ray Diffraction Crystallography. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-16635-8_1

Download citation

DOI : https://doi.org/10.1007/978-3-642-16635-8_1

Published : 09 February 2011

Publisher Name : Springer, Berlin, Heidelberg

Print ISBN : 978-3-642-16634-1

Online ISBN : 978-3-642-16635-8

eBook Packages : Chemistry and Materials Science Chemistry and Material Science (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

Cross-Domain Transfer Learning for Medical Condition Classification from Infant X-ray Images

• Seyoung Park The Webb Schools
• Joe Martin The Webb Schools

Field of artificial intelligence technology has flourished in recent years which led to a creation of diagnosing or assessing diseases using X-ray images. On the downside, these creations focus mostly on adult X-ray images and can not accurately diagnose infant X-ray images. This issue stems from a combination of factors: limited availability of datasets containing infant X-ray and significant variation in these images due to the rapid development of the infant body. Therefore, there is a high demand to develop comprehensive solutions that address these challenges and provide accurate insights. The proposed representation learning-based framework comprises two stages: auto-encoder-based representation learning and transfer learning for diagnosis. The first stage uses adult X-ray images to train the model for improved representation, generating identical reconstructed images. The second stage utilizes pre-trained models to diagnose diseases and predict infant age, enhancing accuracy by accounting for age-related variations in X-ray shapes. This innovative approach represents the first endeavor in unrestricted pediatric X-ray diagnosis, utilizing self-supervised learning for enhanced accuracy. As a result, the comprehensive and extensive experiment allows the proposed method to outperform in comparison to the existing methods. I expect that my research will contribute to the pediatric field of medicine and serve as the foundation of diverging the utility of artificial intelligence.

References or Bibliography

AI Hub (2023, May 1). “Pediatric abdominal x-ray image data”: AI Hub

https://www.aihub.or.kr/aihubdata/data/view.do?currMenu=115&topMenu=100&aihubDataSe=data&dataSetSn=71432

Bhardwaj, Shivam. (2021, Jun 8). “Convolutional Neural Networks : Understand the Basics”: Analytics Vidhya

https://www.analyticsvidhya.com/blog/2021/05/convolutional-neural-networks-understand-the-basics/

Fran, C. (2017). Deep learning with depth wise separable convolutions. In IEEE conference on computer vision and pattern recognition (CVPR). https://doi.org/10.48550/arXiv.1610.02357

He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770-778). https://doi.org/10.48550/arXiv.1512.03385

Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4700-4708). https://doi.org/10.48550/arXiv.1608.06993

Pham, H. H., Nguyen, N. H., Tran, T. T., Nguyen, T. N., & Nguyen, H. Q. (2023). PediCXR: An open, large-scale chest radiograph dataset for interpretation of common thoracic diseases in children. Scientific Data, 10(1), 240.

Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L. C. (2018). Mobilenetv2: Inverted residuals and linear bottlenecks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4510-4520). https://doi.org/10.48550/arXiv.1801.04381

Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556. https://doi.org/10.48550/arXiv.1409.1556

Wang, J., Sun, K., Cheng, T., Jiang, B., Deng, C., Zhao, Y., ... & Xiao, B. (2020). Deep high-resolution representation learning for visual recognition. IEEE transactions on pattern analysis and machine intelligence, 43(10), 3349-3364. https://doi.org/10.48550/arXiv.1908.07919

Xie, S., Girshick, R., Dollár, P., Tu, Z., & He, K. (2017). Aggregated residual transformations for deep neural networks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1492-1500).

https://doi.org/10.48550/arXiv.1611.05431

Zhang, X., Zhou, X., Lin, M., & Sun, J. (2018). Shufflenet: An extremely efficient convolutional neural network for mobile devices. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 6848-6856).

https://doi.org/10.48550/arXiv.1707.01083

How to Cite

• Endnote/Zotero/Mendeley (RIS)

Copyright (c) 2024 Seyoung Park; Joe Martin

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License .

Copyright holder(s) granted JSR a perpetual, non-exclusive license to distriute & display this article.

Announcements

Call for papers: volume 13 issue 4.

If you are a high school student or a recent high school graduate aspiring to publish your research, we are accepting submissions. Submit Your Article Now!

Deadline: 11:59 p.m. August 31, 2024

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Exploring the interplay of dataset size and imbalance on cnn performance in healthcare: using x-rays to identify covid-19 patients.

1. Introduction

1.1. classification biases in cnn due to imbalanced data, 1.2. overcoming data challenges in cnns, 1.3. medical background, 2.1. the classifier, 2.2. convolutional neural network (cnn) architecture.

• First to Fourth Convolutional Layers: Each of these layers is followed by a Max Pooling layer to downsample the feature maps, reducing the spatial dimensions while retaining important features.
• Fifth Convolutional Layer: This layer is followed by a Batch Normalization layer to stabilize and accelerate the training process.
• Flattening Layer: After the convolutional layers, the output is flattened into a one-dimensional array to be fed into the fully connected layers.
• Dense Layer: The flattened output is passed through a dense layer, followed by another Batch Normalization layer to further stabilize training.
• Dropout Layer: A Dropout layer is employed to prevent overfitting by randomly setting a fraction of the input units to zero during training.
• For binary classification (COVID-19 positive vs. negative), the output layer contains two neurons.
• For multi-class classification (e.g., COVID-19, Pneumonia, Normal), the output layer contains three neurons.

2.3. Comparison with Other Networks for COVID-19 Classification

2.4. data description.

• Normal: The dataset consists of 10,192 images, which were further augmented by horizontally inverting the images, resulting in a final count of 10,500 images.
• COVID: The dataset encompasses 3616 images specifically related to COVID-19 cases.
• Lung Opacity: The dataset contains 6012 images of individuals displaying lung opacity.

2.5. Split Data

• “Test 20%” This widely employed approach for training neural networks involves dividing the data in a 70:10:20 ratio. Consequently, 70% of the data is allocated for training, 10% for validation purposes, and 20% for testing. Alternatively, this option can be viewed as an 80:20 split, as 80% of the data is used for assessing and optimizing the model’s performance through training and validation, while the remaining 20% is utilized to evaluate the model’s performance.
• “Test 500”: In addition to the aforementioned option, an alternative approach was adopted, wherein the data were divided into two distinct portions. The first portion was employed for training and validation in an 80:20 ratio, respectively. The second portion was reserved exclusively for testing the model’s performance and included 500 chest X-ray images for each class. The selection of 500 images per class was aimed at achieving a level of accuracy to three decimal points.
• “Fix”: The number of images in the minority class (COVID-19) remained fixed, while the number of images in the majority class was systematically varied during each experiment. The experimentation began with a balanced ratio (50:50) and concluded with a significantly high IR (1:99).
• “Change”: In this approach, the number of images in the minority class was deliberately reduced, while an equivalent increase in the number of images within the majority class was made.

2.6. Imbalance Ratio

2.7. experiments, 4. discussion, cnn performance saturation beyond dataset size, 5. conclusions, supplementary materials, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.

• Rout, N.; Mishra, D.; Mallick, M.K. Handling imbalanced data: A survey. In International Proceedings on Advances in Soft Computing, Intelligent Systems and Applications ; Springer: Singapore, 2018. [ Google Scholar ]
• López, V.; Fernández, A.; García, S.; Palade, V.; Herrera, F. An insight into classification with imbalanced data: Empirical results and current trends on using data intrinsic characteristics. Inf. Sci. 2013 , 250 , 113–141. [ Google Scholar ] [ CrossRef ]
• Li, Q.; Mao, Y. A review of boosting methods for imbalanced data classification. Pattern Anal. Appl. 2014 , 17 , 679–693. [ Google Scholar ] [ CrossRef ]
• Han, C.; Wang, P.; Huang, R.; Cui, L. HCTNet: An experience-guided deep learning network for inter-patient arrhythmia classification on imbalanced dataset. Biomed. Signal Process. Control 2022 , 78 , 103910. [ Google Scholar ] [ CrossRef ]
• Li, D.-C.; Hu, S.C.; Lin, L.-S.; Yeh, C.-W. Detecting representative data and generating synthetic samples to improve learning accuracy with imbalanced data sets. PLoS ONE 2017 , 12 , e181853. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Lee, Z.-J.; Yang, M.-R.; Hwang, B.-J. A Sustainable Approach to Asthma Diagnosis: Classification with Data Augmentation, Feature Selection, and Boosting Algorithm. Diagnostics 2024 , 14 , 723. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Alsalatie, M.; Alquran, H.; Mustafa, W.A.; Zyout, A.; Alqudah, A.M.; Kaifi, R.; Qudsieh, S. A New Weighted Deep Learning Feature Using Particle Swarm and Ant Lion Optimization for Cervical Cancer Diagnosis on Pap Smear Images. Diagnostics 2023 , 13 , 2762. [ Google Scholar ] [ CrossRef ]
• Li, D.; Zheng, C.; Zhao, J.; Liu, Y. Diagnosis of heart failure from imbalance datasets using multi-level classification. Biomed. Signal Process. Control 2023 , 81 , 104538. [ Google Scholar ] [ CrossRef ]
• Lu, W.; Hou, H.; Chu, J. Feature fusion for imbalanced ECG data analysis. Biomed. Signal Process. Control 2018 , 41 , 152–160. [ Google Scholar ] [ CrossRef ]
• Vijayvargiya, A.; Prakash, C.; Kumar, R.; Bansal, S.; Tavares, J.M.R. Human knee abnormality detection from imbalanced sEMG data. Biomed. Signal Process. Control 2021 , 66 , 102406. [ Google Scholar ] [ CrossRef ]
• Rath, A.; Mishra, D.; Panda, G.; Satapathy, S.C. Heart disease detection using deep learning methods from imbalanced. Biomed. Signal Process. Control 2021 , 68 , 102820. [ Google Scholar ] [ CrossRef ]
• Hancer, E.; Samet, M.T.R.; Yıldırım, Z.; Nemati, N. An imbalance-aware nuclei segmentation methodology for H&E stained. Biomed. Signal Process. Control 2023 , 83 , 104720. [ Google Scholar ] [ CrossRef ]
• Arshad, S.; Amjad, T.; Hussain, A.; Qureshi, I.; Abbas, Q. Dermo-Seg: ResNet-UNet Architecture and Hybrid Loss Function for Detection of Differential Patterns to Diagnose Pigmented Skin Lesions. Diagnostics 2023 , 13 , 2924. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Wei, Q.; Dunbrack, R.L., Jr. The role of balanced training and testing data sets for binary classifiers in bioinformatics. PLoS ONE 2013 , 8 , e67863. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kaur, H.; Pannu, H.S.; Malhi, A.K. A systematic review on imbalanced data challenges in machine learning: Applications and solutions. ACM Comput. Surv. (CSUR) 2019 , 52 , 1–36. [ Google Scholar ] [ CrossRef ]
• Mazurowskia, M.A.; Habasa, P.A.; Zuradaa, J.M.; Lob, J.Y.; Bakerb, J.A.; Tourassib, G.D. Training neural network classifiers for medical decision making: The effects of imbalanced datasets on classification performance. Neural Netw. 2008 , 21 , 427–436. [ Google Scholar ] [ CrossRef ]
• Bartosz, K. Learning from imbalanced data: Open challenges and future directions. Prog. Artif. Intell. 2016 , 5 , 221–232. [ Google Scholar ]
• Sun, Y.; Kamel, M.S.; Wang, Y. Boosting for learning multiple classes with imbalanced class distribution. In Proceedings of the Sixth International Conference on Data Mining (ICDM’06), Hong Kong, China, 18–22 December 2006. [ Google Scholar ]
• Pei, X.; Zhao, Y.H.; Chen, L.; Guo, Q.D.Z.; Pan, Y.; Hou, H. Robustness of machine learning to color, size change, normalization, and image enhancement on micrograph datasets with large sample differences. Mater. Des. 2023 , 232 , 112086. [ Google Scholar ] [ CrossRef ]
• Bailly, A.; Blanc, C.; Francis, É.; Guillotin, T.; Jamal, F.; Wakim, B.; Roy, P. Effects of dataset size and interactions on the prediction performance of logistic regression and deep learning models. Comput. Methods Programs Biomed. 2022 , 213 , 106504. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Choi, W.; Lee, S. Performance evaluation of deep learning architectures for load and temperature forecasting under dataset size constraints and seasonality. Energy Build. 2023 , 288 , 113027. [ Google Scholar ] [ CrossRef ]
• He, K.; Zhang, X.; Ren, S.; Sun, J. Identity Mappings in Deep Residual Networks. In Computer Vision—ECCV 2016, Proceedings of the 14th European Conference, Amsterdam, The Netherlands, 11–14 October 2016 ; Springer: Cham, Switzerland, 2016. [ Google Scholar ] [ CrossRef ]
• Huang, G.; Liu, Z.; Maaten, L.V.D.; Weinberger, K.Q. Densely Connected Convolutional Networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 21–26 July 2017; pp. 4700–4708. [ Google Scholar ] [ CrossRef ]
• Hasib, K.M.; Azam, S.; Karim, A.; Marouf, A.A.; Shamrat, F.M.J.M.; Montaha, S.; Yeo, K.C.; Jonkman, M.; Alhajj, R.; Rokne, J.G. CNN-LSTM: Combining CNN and LSTM to Classify Multi-Class Text in Imbalanced News Data. IEEE Access 2023 , 11 , 93048–93063. [ Google Scholar ] [ CrossRef ]
• Londhe, M. Classification of Eye Diseases Using Hybrid CNN-RNN Models. Ph.D. These, National College of Ireland, Dublin, Ireland, 2021. [ Google Scholar ]
• Kotsiantis, S.B.; Pintelas, P.E. Combining bagging and boosting. Int. J. Math. Comput. Sci. 2007 , 1 , 372–381. [ Google Scholar ]
• Kalirane, M. Ensemble Learning in Machine Learning: Stacking, Bagging and Boosting. Analytics Vidhya. 2024. Available online: https://www.analyticsvidhya.com/blog/2023/01/ensemble-learning-methods-bagging-boosting-and-stacking/ (accessed on 15 July 2024).
• Hasib, K.; Iqbal, S.; Shah, F.M.; Mahmud, J.A.; Popel, M.H.; Showrov, I.H.; Ahmed, S.; Rahman, O. A Survey of Methods for Managing the Classification and Solution of Data Imbalance Problem. J. Comput. Sci. 2020 , 16 , 1546–1557. [ Google Scholar ] [ CrossRef ]
• Li, Z.; Chen, Q.; Feng, L.; Rodewald, L.; Xia, Y.; Yu, H.; Zhang, R.; An, Z.; Yin, W.; Chen, W.; et al. Active case finding with case management: The key to tackling the COVID-19 pandemic. The Lancet 2020 , 396 , 63–70. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Gandhi, R.T.; Lynch, J.B.; Del Rio, C. Mild or moderate COVID-19. N. Engl. J. Med. 2020 , 383 , 1757–1766. [ Google Scholar ] [ CrossRef ]
• Berlin, D.A.; Gulick, R.M.; Martinez, F.J. Severe COVID-19. N. Engl. J. Med. 2020 , 383 , 2451–2460. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Peeling, R.W.; Heymann, D.L.; Teo, Y.Y.; Garcia, P.J. Diagnostics for COVID-19: Moving from pandemic response to control. The Lancet 2022 , 399 , 757–768. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Alsharif, W.; Qurashi, A. Effectiveness of COVID-19 diagnosis and management tools: A review. Radiography 2021 , 27 , 682–687. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ebrahimzadeh, S.; Islam, N.; Dawit, H.; Salameh, J.; Kazi, S.; Fabiano, N.; Treanor, L.; Absi, M.; Ahmad, F.; Rooprai, P.; et al. Thoracic imaging tests for the diagnosis of COVID-19. Cochrane Database Syst. Rev. 2022 , 5. [ Google Scholar ] [ CrossRef ]
• Jacobi, A.; Chung, M.; Bernheim, A.; Eber, C. Portable chest X-ray in coronavirus disease-19 (COVID-19): A pictorial review. Clin. Imaging 2020 , 64 , 35–42. [ Google Scholar ] [ CrossRef ]
• Saygılı, A. A new approach for computer-aided detection of coronavirus (COVID-19) from CT and X-ray images using machine learning methods. Appl. Soft Comput. 2021 , 105 , 107323. [ Google Scholar ] [ CrossRef ]
• Nur, A.-A.; Ahsan, M.; Based, M.A.; Haider, J.; Kowalski, M. COVID-19 Detection from Chest X-ray Images Using Feature Fusion and Deep Learning. Sensors 2021 , 21 , 1480. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rahman, T.; Chowdhury, M.; Khandakar, A. COVID-19 Radiography Database. 2021. Available online: https://www.kaggle.com/datasets/tawsifurrahman/covid19-radiography-database (accessed on 13 June 2022).

Click here to enlarge figure

Share and Cite

Davidian, M.; Lahav, A.; Joshua, B.-Z.; Wand, O.; Lurie, Y.; Mark, S. Exploring the Interplay of Dataset Size and Imbalance on CNN Performance in Healthcare: Using X-rays to Identify COVID-19 Patients. Diagnostics 2024 , 14 , 1727. https://doi.org/10.3390/diagnostics14161727

Davidian M, Lahav A, Joshua B-Z, Wand O, Lurie Y, Mark S. Exploring the Interplay of Dataset Size and Imbalance on CNN Performance in Healthcare: Using X-rays to Identify COVID-19 Patients. Diagnostics . 2024; 14(16):1727. https://doi.org/10.3390/diagnostics14161727

Davidian, Moshe, Adi Lahav, Ben-Zion Joshua, Ori Wand, Yotam Lurie, and Shlomo Mark. 2024. "Exploring the Interplay of Dataset Size and Imbalance on CNN Performance in Healthcare: Using X-rays to Identify COVID-19 Patients" Diagnostics 14, no. 16: 1727. https://doi.org/10.3390/diagnostics14161727

Article Metrics

Article access statistics, supplementary material.

ZIP-Document (ZIP, 828 KiB)

Further Information

Mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals