The xerox 125 system was used to make a comparative image quality study between xeroradiography (XR) and industrial type MR film which is approximately equivalent to Kodak type AA film. The comparison of the image quality was made by objective and subjective methods at the visual threshold, since diagnostic details in actual mammograms are often marginably detectable. 1) Dependence of the output density distribution on edge gradient is studied in the spatial and the spatial frequency domain. The test objects employed are aluminum rods having square and round cross-sections of 200 μ which simulate calcification in X-ray absorption. The infinite edge gradient of the square alumiunm rod narrows the width at the image with a wide deletion of the powder outside the boundary, while the alumiunm cylinder increases its image width by attracting extra powder inside the boundary. Density distributions of the alminum test objects are converted to spatial frequency spectra by means of Fourier transformation. The use of a square bar results in an excessive decrease of the output amplitude at low spatial frequencies though greater in amplitude in the intermediate to high spatial frequencies than the cylindrical object. 2) RMS granularity σ(D) is determined by a scanning microdensitometer with a 10×150 μ wide scanning slit adjusted with respect to eye's line spread function. Ganularity of the 125 system is dramatically improved than other xeroradiographic systems which have been seen. It may be due to plastic coating on the transfer paper and smaller toner particle size. However, XR is grainiest among the four imaging systems tested. The density fluctuation of the XR is 4 times greater than MR film. This greater density fluctuation may be a result of the high grain pattern contrast since the effective grain sizes are as small as MR film. 3) The concept of "signal-to-noise ratio" is applied to the imaging systems in order to obtain results from objective methods correlated with subjective results or visibility. The output density distributions of alumiunm cylinders ranging from 100 to 1,000 μ in diameter are taken as the inputs to the visual system, and σ(D), as noise. Physiological contrast D/ΔD min. is also taken into consideration for XR in which the surrounding density of an object varies with the magnitude of the edge-effect. The result reveals that XR provides a higher ratio than MR film for aluminum cylinders smaller than 1,000 μ in diameter. XR differs the most from MR film in the radio at 700 μ diameter or so where XR is considered to show the best visibility of calcifications. 4) Visibility is defined here as the adiliity to detect minute calcification in mammary carcinoma ranging 90〜400 μ in diameter, and comparison of visibility between XR and film is made with following factors ; exposure time, radiographic density and geometrical condition. As a whole, XR is superior to film in visibility at both 30 kVp providing additional detectability for smaller calcification by roughly 20 μ in diameter. When visibility of XR is related to varing exposure time, it is far less sensitive than film, but on the contrary more sensitive than film to the radiographic density. Optimum density for XR is very low as 0.3 were the XR print provides the maximum visibility, while 1.5 for film at 7,000 lux rear illumination. XR is less affected by penumbral effects than film, since the visibility of XR depends much more on the contrast than the shapness for the object larger than 100 μ in diameter. It is concluded that the 125 system satisfied an indespensable necessity of mammography for finding small objects in the breast which a patient or her doctor could not find.
Read full abstract