IMAGE COMPRESSION IN DIGITAL MAMMOGRAPHY - EFFECTS ON COMPUTERIZED DETECTION OF SUBTLE MICROCALCIFICATIONS

Our previous receiver operating characteristic (ROC) study indicated t hat the detection accuracy of microcalcifications by radiologists is s ignificantly reduced if mammograms are digitized at 0.1 mmx0.1 mm. Our recent study also showed that detection accuracy by computer decrease s as the pixel size increases from 0.035 mmx0.035 mm. It is evident th at very large matrix sizes have to be used for digitizing mammograms i n order to preserve the information in the image. Efficient compressio n techniques will be needed to facilitate communication and archiving of digital mammograms. In this study, we evaluated two compression tec hniques: full frame discrete cosine transform (DCT) with entropy codin g and Laplacian pyramid hierarchical coding (LPHC). The dependence of their efficiency on the compression parameters was investigated. The t echniques were compared in terms of the trade-off between the bit rate and the detection accuracy of subtle microcalcifications by an automa ted detection algorithm. The mean-square errors in the reconstructed i mages were determined and the visual quality of the error images was e xamined. It was found that with the LPHC method, the highest compressi on ratio achieved without a significant degradation in the detectabili ty was 3.6:1. The full frame DCT method with entropy coding provided a higher compression efficiency of 9.6:1 at comparable detection accura cy. The mean-square errors did not correlate with the detection accura cy of the microcalcifications. This study demonstrated the importance of determining the quality of the decompressed images by the specific requirements of the task for which the decompressed images are to be u sed. Further investigation is needed for selection of optimal compress ion technique for digital mammograms. (C) 1996 American Association of Physicists in Medicine.