A Comprehensive Guide to Radiographic Sciences and Technology. Euclid Seeram

FPDR or DR as it is sometimes referred to, DF, DM, digital radiographic tomosynthesis (DRT), digital breast tomosynthesis (DBT), and last but not least CT. Additionally, since all of the above‐mentioned modalities include image processing using computers, the concepts of Digital Image Processing will be reviewed since it has become an essential tool for technologists, radiologists, and medical physicists working in a digital radiology department.

These imaging modalities include specific physics concepts that must be understood for optimum results. For example, CR is based on the use of photostimulable phosphors (PSP) which are based on the physical principle of photostimulable luminescence (PSL). An example of one such phosphor is barium fluoro halide (BaFX) where the halide (X) can be chlorine (Cl), bromine (Br), iodine (I), or a mixture of them. When the PSP imaging plate (IP) is exposed to x‐rays, electrons are moved from the ground state (valence band) to a higher energy level (conducting band) and are trapped there until the PSP plate is exposed to a laser light and subsequently the electrons in the higher energy state return to their ground state, thus emitting a bluish‐purple light referred to as PSL.

      The detectors used in FPDR are based on semiconductor physics. Examples of two such common detectors used in DR are indirect digital detectors which use amorphous silicon photoconductor coupled to an x‐ray scintillator (cesium iodide for example) and direct digital detectors which use amorphous selenium photoconductor. While the former detector converts x‐rays to light which falls upon the silicon photoconductor to produce electrical signals, the latter detector converts x‐rays directly into electrical signals. The other digital imaging modality detectors are based on photoconductor physics.

      The imaging modalities listed above convert radiation attenuated by the patient and falling on the digital detector to digital data. This is necessary since computers are used to process these data through popular digital image processing operations. These operations have become commonplace and must be fully understood for effective use in clinical practice. One such tool is the concept of windowing, where the image brightness and contrast can be changed by the operator to suit the viewing needs of the human interpreter. Furthermore, other digital image processing tools that are vital in DBT and CT image reconstruction algorithms. These algorithms have evolved from the filtered back projection (FBP) algorithm to more complex algorithms such as iterative reconstruction (IR) algorithms. These algorithms play an important role in building up an image from data collected through 360° around the patient in CT, for example. Today, IR algorithms are now used by all CT vendors.

      Radiographic exposure technique

      Radiographic exposure technique refers to the use of exposure factors coupled with other elements on the x‐ray control panel, selected by the technologist to produce diagnostic images. Exposure factors include the kilovoltage (kV), the milliamperes (mA), and exposure time (s) and the selection of the appropriate source‐to‐image receptor distance (SID). Furthermore, the proper positioning of the patient and image receptor, tube alignment with the image receptor, use of appropriate filtration and collimation, and patient instructions, are all the other elements that play an important role during the radiographic examination.

      Image quality considerations

Image quality is a significant goal of radiographic imaging modalities. The attenuated radiation data from the patient are used to create images that are used for diagnostic interpretation by a human observer. There are at least five important descriptors of digital image quality and these include spatial resolution, contrast resolution, noise, detective quantum efficiency (DQE), and image artifacts. While spatial resolution addresses the sharpness of images, and is related to the size of the pixels (picture elements) in an image, contrast resolution or density resolution deals with the ability of the imaging system to demonstrate differences in tissue contra, and is linked to the bit depth, that is the range of gray levels per pixel. Noise, on the other hand, depends on the number of x‐ray photons used to create the image. While fewer photons (low exposure technique factors) will result in more noise (grainy appearance), more photons (higher exposure technique factors) will create a better image (less noisy image), but at the expense of dose. Another digital image quality descriptor is the DQE, which is a measure of the efficiency and fidelity with which the detector can convert an input exposure into a useful output image. Finally, digital images are not free of artifacts. These are features seen on the image that are not present in the patient, and can pose challenges for the human observer in detecting fact from artifact.

      Computed tomography – physics and instrumentation

      This section will present a broad overview of the essential elements of the Physics and Instrumentation of Computed Tomography. One of the major advantages of CT is that it provides improved contrast resolution compared to radiography and for this reason, it has proven to be worthy of further developments in imaging soft tissues of the human body. It is important to note, however, that magnetic resonance imaging (MRI) has superior contrast resolution compared to all other imaging modalities, such as radiography, nuclear medicine, and diagnostic medical sonography.

      CT is a sectional imaging technique that produces direct cross‐sectional digital images referred to as transverse axial images which has been referred to as planar sections that are perpendicular to the long axis of the patient. The word “computed” implies that a computer is used to process and reconstruct x‐ray transmission data collected from the patient. The CT scanner has evolved from single‐slice CT scanners (SSCT) to multi‐slice CT scanners (MSCT). State‐of‐the‐art CT scanners are now MSCT scanners capable of a wide range of applications. The increasing use is that CT in clinical practice has led to increasing doses to the patient and a well‐documented fact is that CT delivered the highest collective dose in the United States compared to other medical imaging modalities.

      Two individuals shared the Nobel Prize in Medicine and Physiology in 1979 for their development of the CT scanner. These include Godfrey Newbold Hounsfield in the United Kingdom (UK) who invented the first clinically useful scanner, and Allan Cormack, a physicist at Tufts University in Massachusetts.

      CT is a multidisciplinary technology and has its roots in physics, mathematics, engineering, and computer science. The CT process consists of at least three major system components that are used to produce the CT image; the data acquisition system; the computer system; and the image display, storage, and communication systems.

      Data acquisition means that radiation attenuation data are collected from the patient during the scanning. In this respect, an x‐ray tube coupled to special electronic detectors rotate around the patient to collect and measure attenuation readings as the x‐ray beam passes through the patient.

      The attenuation is according to Beer–Lambert's law:

where I is the transmitted x‐ray beam intensity, I_o is the original x‐ray beam intensity, e represents Euler's constant, μ is the linear attenuation coefficient, and Δx is the finite thickness of the section. In CT, the system calculates all μs for all structures seen on the image. Special detectors and detector electronics are used to calculate the attenuation data and convert them into integers (0, a positive number, or a negative number) referred to as CT numbers using an image reconstruction algorithm to build up the image in numerical format. The CT numbers (numerical image format) are converted into a gray‐scale image for display on a monitor for the observer to interpret.

      The CT numbers are calculated using the following relationship:

      where K represents a scaling factor. In general, K is equal to 1000. When Hounsfield invented the scanner, K was equal to 500.

      The technology aspects