Comparisons of different contrast resolution effects on a computer-aided detection system intended to cluster microcalcifications detected in dense breast images.

Clustered microcalcifications, which are frequently an important signal of possible cancer, are usually hidden in dense breast images, adding more difficulty in mammogram medical analysis. In this work we evaluate the performance of a previously developed computer-aided detection scheme, modified for application to dense breast images. The main focus of this investigation was on the effect of different contrast resolutions on the processing performance. We have processed dense breast images digitized with 8 and 12 bits to evaluate the performance of this computer-aided detection scheme with different contrast resolutions. As expected, for most of the 12-bit images, the number of detected signals was greater or at least equal to that of the 8-bit images.

Investigations on the effect of different characteristics of images sets on the performance of a processing scheme for microcalcifications detection in digital mammograms.

The performance of a computer-aided diagnosis (CAD) scheme is closely dependent on the database used for its development and tests. The scheme sensitivity can be reduced by 15% to 25%, with only 20% of changes in the database cases. Previously, we have developed a processing scheme in order to detect clustered microcalcifications in digital mammograms, and we have tested such a procedure with two different databases. Further evaluations in developing a CAD scheme for mammography have indicated the need for more extensive investigation on the effects resulting from different characteristics of the images bank used for tests. Therefore, this work reports some results regarding such an investigation, with a further discussion over characteristics that can affect the performance of a CAD scheme.