Digital XRay Mammography

In screen-film mammography, a phosphor screen in a light-tight cassette absorbs a fraction of the incident x rays. This fraction, typically 60 to 80 percent is known as the quantum efficiency. The phosphor also converts the energy to light and has a certain conversion efficiency for this process. The light is coupled to a sheet of photographic film by direct contact of the screen and film within the cassette. The signal is recorded in the form of a latent photographic image on the film. This is developed by chemical processing to produce a pattern of optical density on the film, which is then viewed by transillumination. The film itself is both the recording and display device and, in addition, is the archival record of the examination.

In digital mammography, the image acquisition and display operations are separated. The image is acquired by a detector which converts the x-ray signal into electronic form, and then it is digitized or quantized into one of two to the nth power (2n) intensity levels. Typically, n, the number of bits of digitization, is 12 or 14, giving 4,096 or 16,384 image signal levels. The digital image is also sampled spatially (i.e., either the detector surface is composed of separate x-ray sensitive elements or else the output signal from a continuous detector is broken up into discrete elements each representing the signal from a small area at the detector's entrance).

Screen-film mammography has inherent physical limitations which reduce its effectiveness.

• The film gradient needed for high contrast must be balanced against the need for wide latitude. This is illustrated in Figure 3.21, which is a characteristic curve for a mammo-graphic screen-film combination. The gradient of the curve falls off both at low and high exposures, resulting in a loss of contrast in those regions. If the film gradient is increased, the range of exposures between the minimum and maximum optical densities on the film decreases further.

• Detection of microcalcifications and their portrayal with clarity of the margins of breast masses are reduced due to the presence of film noise and screen blur in the displayed image.

• Film-processing artifacts occasionally degrade the mammo-graphic image.

• The day-to-day variability in performance of automated film processors can produce suboptimal image quality.

In digital mammography, the processes of image acquisition and display are separated so that each can be optimized independently. The image is stored as a matrix of numbers, where each number represents for a specific small square or "pixel" in the image, the number of x rays reaching that point after having been transmitted by the breast (Yaffe, 1992).

3.3.1 Digital Imaging Detectors

Detectors for digital mammography can be designed to have a linear response to x rays over a very wide range of exposures (Figure 3.22). After the image data are recorded, it is then possible to apply a transformation to display the image on a high-resolution

0 0

Post a comment