Region Prefiltering

Feature extraction is computationally expensive. A common strategy [6, 7,36] to reduce the number of regions considered for false-positive reduction is achieved by applying a size test. By eliminating suspicious regions smaller than a predefined threshold 71area, the number of false-positive regions can be reduced. For the expert radiologist interpreting a film mammogram during screening, it is common to disregard any suspicious ROI less than 8 mm in diameter [37]. In mammographic CAD with computer automation, the size threshold is reduced and a common value for Tarea is the number of pixels corresponding to an area of 16 mm2 [6, 7, 36]. In the adaptive knowledge-based model, the area threshold is set at 19.5 mm2 corresponding to a region diameter of 5 mm for all breast type groupings. The DDSM used in this evaluation are digitized such that each pixel is 50 ^m. Following subsampling by a factor of four, an area threshold of 19.5 mm2 is equivalent to Tarea = 122 pixels, thus any suspicious region following segmentation with an area less than this value is marked as normal.

0 0

Post a comment