There are a large number of factors that make a digital image. Many can be influenced by the radiologic technologist. Some factors are influenced by the design of the system being used. No single factor is more important than the Histogram. Histograms are the basis of which many other factors are derived.
A histogram is a distribution frequency of digital values in an image. Histograms can be found in all forms of photography, mainly digital imagery.
In digital imaging there are numerous density values that represent various tissue densities. As the imaging system recognizes all these values, it constructs a gray-scale histogram of them representing the anatomical characteristics of the imaged part. Thus all PA chest histograms will be similar, all lateral chest histograms will be similar, all pelvis histograms will be similar, etc.
Before an imaging system is able to process the digital image, the radiographer selects a processing algorithm by selecting the anatomical part and particular projection on the control panel at the workstation. The workstation then matches that information with a particular look-up- table (LUT). The system then uses the inputted information from the user and the histogram derived from the image its self to process the image into a diagnostic quality. Because all post processing is based upon the original histogram and user input, it is extremely important to reduce factors that may skew a histogram such as metal on the patient.
A look-up-table (LUT) is a data set of values and algorithms that a particular combination of histogram values and input information is compared to in an effort to apply the desired post processing algorithms. In essence, the LUT is used as a reference to evaluate the raw information and correct density values.
This processing function selects all pixels (each with its own specific gray value) are changed to a new gray value. The resultant image will have the appropriate appearance in brightness and contrast.
Histogram appearance can be affected by a number of things. Accuracy in positioning and centering can have a significant effect on histogram appearance. Because of poor collimation the average exposure level has changed, as well as the exposure's latitude; these changes will be reflected in the images' exposure index. Other factors affecting histogram appearance include selection of the correct processing algorithm (e.g., chest vs. femur vs. cervical spine), changes in scatter, SID, OID, foreign densities (metal), and collimation. In short, anything that affects scatter and/or dose. One other factor is delay in processing from time of exposure. Processing delay can result in fading of the image. Normal examination times and short delays between projections will generally not be a problem.
Brightness is the intensity of light that represents the individual pixels in the image on the monitor. In digital imaging, the term brightness replaces the film-based term density.
Digital imaging systems are designed to electronically display the optimal image brightness under a wide range of exposure factors. Brightness is controlled by the processing software through the application of predetermined digital processing algorithms. This is termed "Automatic Brightness Control" or ABC and is similar to the ABC in a fluoroscopic unit only it is not mechanical but a software feature in digital imaging. Unlike the relationship between mAs and density in film-screen imaging, changes in mAs does not have a controlling effect directly on digital image brightness. This is because the powerful post processing algorithms used in digital imaging and is part of what makes digital imaging so forgiving when compared to film. In addition, the user can adjust the brightness of the digital image after exposure.
In digital imaging, contrast is the difference in brightness between light and dark areas of an image. Contrast resolution refers to an imaging system's ability to distinguish between similar tissues (densities).
Digital imaging systems are designed to electronically display optimal image contrast, under a wide range of exposure factors. In film-screen imaging, kV is the controlling factor for image contrast; however, in digital imaging, image contrast is affected by the application of predetermined algorithms. Therefore, the user can adjust the contrast of the digital image after exposure. Typically it is the Radiologist's preference on what contrast algorithms are used for each type of anatomy.
Resolution in digital imaging is the recorded sharpness or detail of structures on the image, the same as described for film-screen imaging.
Resolution in a digital image is a combination of the traditional film-screen factors (foal spot size, geometric factors, and motion) and the acquisition pixel size that is inherent to the digital detector. Resolution in digital imaging is typically measured in microns rather than line pairs. The current range for digital general radiographic imaging is from 100 to 200 microns (approximately 2.5 to 5 line pairs per mm). In addition to pixel size, resolution is also controlled by the display capabilities (pixels) of the monitor (pixel pitch).
Distortion is the misrepresentation of object size or shape as projected onto radiographic recording media, just as for film-screen imaging. The factors that affect distortion (SID, OID, and central ray alignment) are therefore the same for film-screen imaging and digital imaging.
Exposure index(s) in digital imaging is a numeric value that is representative of the exposure received by the PSP screen or detector. It is calculated from the effect of mAs, kV, total detector area irradiated, and objects exposed. For example, Fuji uses the Sensitivity or "S" number which is inversely proportional to the radiation striking the detector. For example, if the range for an acceptable "S" number is 150 to 200, an "S" value higher than 250 would indicate underexposure and a value lower than 150 would indicate overexposure. Another exposure index is the Exposure Indicator (EI) by Kodak and is directly proportional to the radiation striking the IR as determined by logarithmic calculations.For example, if an acceptable exposure index is typically 1700-2300, an index value lower than 1700 would indicate underexposure, whereas an index value higher than 2300 would indicate overexposure.
It is important to note a few things when dealing with exposure index(s):
Despite the wide latitude regarding exposure factors with digital imaging, the radiographer must make sure that the exposure factors used were adequate. Checking the exposure index is a helpful tool in verifying that optimal quality digital radiographic images were obtained with the least possible dose to the patient. Even if the exposure index is outside the recommended range the image may still appear acceptable when viewed on the monitor of the radiographer's workstation.
Noise is a random disturbance that obscures or reduces clarity which translates into a grainy or mottled appearance of the image. One way to describe noise is by using the concept of signal-to-noise ratio (SNR).
The number of x-ray photons that strike the detector is considered the "signal" and the factors that negatively affect the final image are classified as "noise." A high SNR is desirable, where the signal is greater than the noise. A low SNR is undesirable, where a low signal with high noise obscures soft-tissue detail and demonstrates a grainy or mottled image. Generally speaking, SNR increases as mAs increases, but so does patient dose. Although a high SNR is favorable, radiographers must ensure that exposure factors used are not beyond what is required for the projection so as not to overexpose the patient.
When insufficient mAs is selected for a projection, the detector does not receive the appropriate amount of radiation, resulting in a low SNR and a noisy image. This mottle may not be readily visible on the lower-resolution monitor of the radiographer's workstation, but the exposure index can determine this. Scatter radiation is a potential source of noise that can be controlled by the use of grids and correct collimation. System noise, artifacts, and unevenness in an image are determined by the noise inherent in the imaging plate (structure noise) and reader system (quantum noise).
It is important to remember that the monitor the radiographer uses to view the image is typically of lower resolution than the radiologist's reporting workstation. The radiographer's workstation is intended to allow verification of positioning and general image quality; however, this image is not of diagnostic quality. The monitor of a radiologist's reporting workstation typically provides superior spatial and contrast resolution due to increased display matrix with smaller pixels and superior brightness characteristics.
Digital images are stored in pixel form, which is a two-dimensional "picture element" measured in the "XY" direction. The third dimension, "Z" direction, in the matrix of pixels is the depth that is referred to as the voxel (volume element). Each pixel demonstrates a single shade of gray when viewed on a monitor and it is representative of the anatomic structure. The range of possible shades of gray demonstrated is related to the pixel's bit depth, which is determined by the manufacturer. The greater the bit depth of a system, the greater the contrast resolution.
Since computer theory is based on the binary system, a 14-bit system, for example, is represented as 214; the 14-bit-deep pixel could represent any one of 16,384 possible shades of gray, from black to white. Bit depth is determined by the manufacturer's system design and is closely related to the imaging procedures the equipment is designed for.
An electronic digital image is formed by a matrix of pixels in rows and columns. A matrix having 512 pixels in each row and column is a 512 x 512 matrix. Fewer and larger pixels result in a poor-resolution "pixely" image, that is, one in which you can actually see the individual pixel boxes.
There are two pixel sizes in medical imaging. These are acquisition pixel size, which is the minimum size that is inherent to the acquisition system, and display pixel size, which is the minimum pixel size that can be displayed on a monitor. A general radiography acquisition matrix may be as high as 3000 x 3000 pixels—over 9 million pixels (9 megapixels) in a 17 x 17-inch (43 x 43 cm) image.