September 9, 2009 - Parascript will present at the 95th Scientific Assembly and Annual Meeting McCormick Place, Chicago, Ill., Nov. 29 – Dec. 4, 2009, in booth 2215, AccuDetect, its next generation computer-aided detection algorithms designed for the early detection of breast lesions.
This solution is based on proprietary image analysis and pattern recognition technology. Parascript image analysis algorithms are designed to detect calcifications and their characteristics such as cluster density (number of calcifications in a cluster), shape of the calcifications, and other important features necessary to determine lesion significance. For masses, the image analysis algorithms look at degree of spiculation, lesion shape, contrast to surrounding tissue, texture of lesion interior, edge texture, and other important characteristics that determine lesion significance.
Biologically significant features are augmented by features automatically extracted by proprietary learning algorithms. The integration of several complementary approaches enables the use of sophisticated voting methods to achieve highest sensitivity and lowest false positive rates. AccuDetect 3.0, introduced at RSNA 2009 is even more robust with improved performance and increased potential for lowering false-positive rates in detecting suspicious lesions on mammograms. Available for OEM customers interested in reducing false-positive rates of existing CAD systems, AccuDetect is intended to assist radiologists in the early detection of breast cancer during film-based or digital mammography exams.
For more information: www.parascript.com