Digital Image Processing (P1) 2024 PDF

Summary

This document provides a comprehensive overview of digital image processing, focusing on characteristics such as pixel size, bit depth, the relationship among pixel size, FOV, and matrix size, dynamic range, and other relevant parameters.. The document is intended for use in medical imaging.

Full Transcript

Digital Image Processing (P1) Lecturer : Hanan Alayfei M.Sc. HI BSc. E-mail: halyafei@ksu.edu.sa DIGITAL IMAGE CHARACTERISTICS In digital imaging, the latent image is stored as digital data and must be processed by a computer for vi...

Digital Image Processing (P1) Lecturer : Hanan Alayfei M.Sc. HI BSc. E-mail: halyafei@ksu.edu.sa DIGITAL IMAGE CHARACTERISTICS In digital imaging, the latent image is stored as digital data and must be processed by a computer for viewing on a display monitor. Two types of digital radiographic systems are in common use today: computed radiography (CR) and direct radiography (DR). In both CR & DR, the computer can manipulate the radiographic image in various ways after the image has been digitally created. Computers can also perform various postprocessing image manipulations to further improve the visibility of the anatomic region. Digital images are composed of numerical data that can be easily manipulated by a computer. A digital image is recorded as a matrix or combination of rows and columns (array) of small “picture elements” called pixels. The size of a pixel is measured in microns (100 microns = 0.1 mm). Each pixel is recorded as a single numerical value, which is represented as a single brightness level on a display monitor. The location of the pixel within the image matrix corresponds to an area within the patient or volume of tissue. Given the dimensions of an anatomic area, or !eld of view (FOV), a matrix size of 1024 X 1024 has 1,048,576 individual pixels; a matrix size of 2048 X 2048 has 4,194,304 pixels. Digital image quality is improved with a larger matrix size that includes a greater number of smaller pixels. Although image quality is improved for a larger matrix size and smaller pixels, computer processing time, network transmission time, and digital storage space increase as the matrix size increases. There is a relationship among pixel size, FOV, and matrix size, as demonstrated in the following formula: Bit depth: Representative of the number of shades of gray that can be demonstrated by each pixel. Bit depth is determined by the manufacturer and is based on the imaging procedures for which the equipment is required. The numerical value assigned to each pixel is determined by the relative attenuation of x-rays passing through the corresponding volume of tissue. Pixels representing highly attenuating tissues (increased absorption) such as bone are usually assigned a low value for higher brightness than pixels representing tissues of low x-ray attenuation (decreased absorption) Each pixel also has a bit depth, or number of bits (Box 4-2), that determines the amount of precision in digitizing the analog signal and therefore the number of shades of gray that can be displayed in the image. Bit depth is determined by an analog-to-digital converter, which is an integral component of every digital imaging system. Because the binary system is used, bit depth is expressed as 2 to the power of n, or the number of bits (2n). A larger bit depth allows a greater number of shades of gray to be displayed on a computer monitor. 12 bit depth (212) can display 4096 shades of gray, 14 bit depth can display 16,384 shades of gray, 16 bit depth can display 65,536 shades of gray. A system that can digitize and display a greater number of shades of gray has better contrast resolution. An image with increased contrast resolution increases the visibility of anatomic details and the ability to distinguish among small anatomic areas of interest. An image consisting of a greater number of pixels per unit area, or pixel density, provides improved spatial resolution. In addition to its size, the pixel spacing or distance measured from the center of a pixel to an adjacent pixel determines the pixel pitch (Figure 4-4). Smaller-sized pixels will have decreased pixel pitch and improved spatial resolution. Spatial Frequency and Spatial Resolution Spatial resolution in digital imaging is primarily limited to the size of the pixel. Anatomic details are composed of large and small objects and radiographic images display those details as variations from white-to- black brightness levels. Small objects have higher spatial frequency and large objects have lower spatial frequency. Spatial frequency can be defined by the unit of line pairs per millimeter (lp/mm). A resolution test pattern is a device used to record and measure line pairs (Figure 4-5). A resolution test pattern is a device used to record and measure line pairs (Figure 4-5). An imaging system that can resolve a greater number of line pairs per millimeter (higher spatial frequency) has increased spatial resolution (Figure 4-6). In digital imaging systems, the ability to resolve or demonstrate a specific spatial frequency is directly impacted by the size of the pixel. Dynamic Range The dynamic range of a digital imaging system refers to the ability of a detector to accurately capture the range of photon intensities that exit the patient. The digital IRs have a much wider dynamic range (Figure 4-21). This wide dynamic range means that a small degree of underexposure or overexposure would still result in acceptable image quality. A numerical value (digital data) is assigned to the pixel that represents the attenuation characteristics of that volume of tissue. If optimal exposure techniques are not used, the image rescaling that occurs during this processing step can produce images with the appropriate brightness levels. Lower-than-necessary x-ray exposures can be detected and processed, image quality suffers because there is insufficient exposure to the IR and quantum noise results. Detective Quantum Efficiency Detective quantum efficiency (DQE) is a measurement of the efficiency of an image receptor in converting the x-ray exposure it receives to a quality radiographic image. the higher the DQE of a system, the lower the radiation exposure required to produce a quality image, thereby decreasing patient exposure. The system’s DQE value is impacted by both the type of material used in the image receptor to capture the exit radiation and the energy of the x-ray. The DQE is higher for DR when compared to CR. Signal-to-Noise Ratio Signal-to-noise ratio (SNR) is a method of describing the strength of the radiation exposure compared with the amount of noise apparent in a digital image. The photon intensities are converted to an electronic signal that is digitized by the analog-to-digital converter (ADC), The term signal refers to the strength or amount of radiation exposure captured by the IR to create the image. Increasing the SNR improves the quality of the digital image; this means that the strength of the signal is high in comparison with the amount of noise, so image quality is improved. Quantum noise results when there are too few x-ray photons captured by the IR to create a latent image. Quantum noise results when there are too few x-ray photons captured by the IR to create a latent image. Both quantum noise and the electronics are sources of noise that capture, process, and display the digital image. The ability to visualize anatomic tissues is affected by the SNR. Histogram Analysis Histogram analysis is an image-processing technique commonly used to identify the edges of an image and assess the raw data prior to image display. The computer first creates a histogram (or graphic representation of a data set) of the image (Figure 4-23). A data set includes all the pixel values that represent the image before edge detection and rescaling. This graph represents the number of digital pixel values versus the relative prevalence of the pixel values in the image. The x-axis represents the amount of exposure, and the y-axis represents the incidence of pixels for each exposure level. The computer analyzes the histogram using processing algorithms and compares it with a pre-established histogram specific to the anatomic part being imaged. This process is called histogram analysis. The computer software has stored histogram models, each having a shape characteristic of the selected anatomic region and projection. These stored histogram models have values of interest (VOI), which determine the range of the histogram data set that should be included in the displayed image. In CR imaging, the entire imaging plate is scanned to extract the image from the photostimulable phosphor. The computer identifies the exposure !eld and the edges of the image, and all exposure data outside this !eld are excluded from the histogram. All four edges of a collimated field should be recognized. If at least three edges are not identified, all data, including raw exposure or scatter outside the field, may be included in the histogram, resulting in a histogram- analysis error. Histogram-analysis errors are less likely to occur with DR IRs compared with CR IRs because the image data are extracted only from the exposed detectors. Automatic Rescaling Histogram analysis is also employed to maintain consistent image brightness despite overexposure or underexposure of the IR. This procedure is known as automatic rescaling. The computer rescales the image on the basis of the comparison of the histograms, which is actually a process of mapping the grayscale to the VOI to present a specific display of brightness (Figure 4-25). Although automatic rescaling is a convenient feature, radiographers should be aware that rescaling errors occur for a variety of reasons and can result in poor-quality digital images. Lookup Tables Following histogram analysis, lookup tables provide a method of altering the image to change the display of the digital image in various ways. Because digital IRs have a linear exposure response and a very wide dynamic range, raw data images exhibit low contrast and must be altered to improve the visibility of anatomic structures. Lookup tables provide the means to alter the brightness and grayscale of the digital image using computer algorithms. They are sometimes used to reverse or invert image grayscale. Figure 4-26 visually compares pixel values of the original image with those of a processed image. If the image is not altered, the graph would be a straight line. If the original image is altered, the original pixel values would be different in the processed image and the graph would no longer be a straight line but might resemble a characteristic curve for radiographic film (Figure 4-27). For example, each pixel value could be altered to display the digital image with a change in contrast. New pixel values would be calculated that result in the image being displayed with higher contrast (Figure 4-28). Figure 4-29 shows the original image, the graph following changes in the pixel values, and the processed higher-contrast image. Lookup tables provide a method of processing digital images in order to change the displayed brightness and contrast required for each anatomic area (Figure 4-30).! THANK YOU

Use Quizgecko on...
Browser
Browser