Microcalcification detectability for four mammographic detectors: flat-panel, CCD, CR, and screen/film)

Abstract

Amorphous silicon/cesium iodide (a-Si:H/CsI:Tl) flat-panel (FP)-based full-field digital mammography systems have recently become commercially available for clinical use. Some investigations on physical properties and imaging characteristics of these types of detectors have been conducted and reported. In this perception study, a phantom containing simulated microcalcifications (microCs) of various sizes was imaged with four detector systems: a FP system, a small field-of-view charge coupled device (CCD) system, a high resolution computed radiography (CR) system, and a conventional mammography screen/film (SF) system. The images were reviewed by mammographers as well as nonradiologist participants. Scores reflecting confidence ratings were given and recorded for each detection task. The results were used to determine the average confidence-rating scores for the four imaging systems. Receiver operating characteristics (ROC) analysis was also performed to evaluate and compare the overall detection accuracy for the four detector systems. For calcifications of 125-140 microm in size, the FP system was found to have the best performance with the highest confidence-rating scores and the greatest detection accuracy (Az = 0.9) in the ROC analysis. The SF system was ranked second while the CCD system outperformed the CR system. The p values obtained by applying a Student t-test to the results of the ROC analysis indicate that the differences between any two systems are statistically significant (p<0.005). Differences in microC detectability for the large (150-160 microm) and small (112-125 microm) size microC groups showed a wider range of p values (not all p values are smaller than 0.005, ranging from 0.6 to <0.001) compared to the p values obtained for the medium (125-140 microm) size microC group. Using the p values to assess the statistical significance, the use of the average confidence-rating scores was not as significant as the use of the ROC analysis p value for p value.