Abstract: A system, method, and computer program product for computer-aided detection of suspicious lesions in digital mammograms is described, wherein single-view feature vectors from a first digital mammogram are processed in a classification algorithm along with information computed from a plurality of related digital mammograms to assign an overall probability of suspiciousness to potentially suspicious lesions in the first digital mammogram. In one preferred embodiment, a greater probability of suspiciousness is determined where there are similar corresponding lesions in the first digital mammogram and in an alternate digital mammogram view of the same breast. In another preferred embodiment, a lesser probability of suspiciousness is found where there are symmetric lesions or structures located in the first digital mammogram and a digital mammogram of the opposite breast.

Abstract: A method and apparatus for the automatic segmentation of the pectoral muscle boundary in a digital mammogram is disclosed. Parameters of the pectoral boundary line are determined by calculating gradient magnitudes of the digital mammogram in a region of interest, accumulating the gradient magnitudes into a parameter plane according to a Hough transform, and using information in the parameter plane for identifying the pectoral boundary. Prior to being accumulated in the parameter plane, the gradient magnitudes are weighted by a monotonic, multi-bit weighting function designed to exhibit larger variations near more densely populated levels of gradient magnitudes, and designed to exhibit smaller variations near less densely populated levels of gradient magnitudes, for allowing greater robustness against variations in gradient magnitudes among different digital mammograms.

Abstract: A method an system for using a local attention threshold to aid in the detection of clustered abnormalities in digitized medical images is disclosed. The local attention threshold is applied to locate spots within a predetermined distance from previously identified spots. More specifically, seed pixels are identified by applying a first seed threshold function to the output of a shift-invariant neural network and adaptive threshold. The seed pixels are then segmented into spots by applying a segmentation threshold function to each seed pixel. False-positive spots are removed using various techniques. Additional seed pixels are then identified by applying a local attention threshold to pixels within a predetermined distance to previously identified spots. The local attention threshold disclosed is less selective for pixels which are closer to the nearest spot than for pixels which are further from the nearest spot.

Abstract: A user interface for facilitating the input of films into a computer-aided diagnosis system is disclosed. The user interface includes a scanner which receives and scans the film-based images, and a film feeder which holds and transports them to the scanner. A touch sensitive display screen is provided to display status information to and receive instructions from an operator. The display screen displays miniature images of the films having color-coded borders so that the operator monitors the processing of the films. The interface also includes a bar code reader and allow the operator to conduct procedures for testing the scanner.

Abstract: A method and apparatus for the fast detection of spiculated lesions in a digital mammogram, the method for use in a computer aided diagnosis system for assisting a radiologist in identifying and recognizing the spiculations among a multiplicity of lines corresponding to standard fibrous breast tissue. A line and direction image is created from a digital mammogram, and a region of potential intersection for substantially every pixel in the digital mammogram image is determined. The region of potential intersection for each pixel is a predetermined pattern, such as a high aspect ratio rectangle or trapezoid, positioned around the pixel and rotated in a direction corresponding to direction information for that pixel. The regions of potential intersection are accumulated among the pixels to produce a cumulative array, and information in the cumulative array is processed for identifying spiculations in the digital mammogram.