University of Pittsburgh

Bin Zheng

David Gur

Yuan-Hsiang Chang

Walter F Good

Glenn S Maitz

Xiao Hui Wang

Yuang-Hsiang Chang

Abnormal regions in living tissue are detected by obtaining images from different views of the living tissue; performing single-image CAD of each image to determine suspected abnormal regions depicted in the image; and combining measurements of the suspected abnormal regions in each image to determine whether a suspected abnormal region is an abnormal region. The living tissue may be a human breast and the abnormal region may be a mass in the breast. Ipsilateral mammographic views of the breast, a craniocaudal view, and a mediolateral oblique view may be used. Features which are relatively invariant or behave predictably with respect to breast compression are extracted using the single-image CAD and then combined.

Multi-view computer-assisted diagnosis

A computerized method of detecting regions of interest in a digital image optimizes and adapts a computer aided scheme for detecting regions of interest in images. The optimization is based on global image characteristics. For each image in a database of images having known regions of interest, global image features are measured and an image characteristic index is established based on these global image features. All the images in the database are divided into a number of image groups based on the image characteristic index of each image in the database and the CAD scheme is optimized for each image group. Once the CAD scheme is optimized, to process a digital image, an image characteristics based classification criteria is established for that image, and then global image features of the digitized image are determined. The digitized image is then assigned an image characteristics rating based on the determined global image features, and the image is assigned to an image group based on the image rating. Then regions of interest depicted in the image are determined using a detection scheme adapted for the assigned image group.

Image quality based adaptive optimization of computer aided detection schemes

A method and apparatus for detecting abnormal regions in living tissue depicted in a digital radiograph includes identifying suspected abnormal regions depicted in the radiograph and then, for each identified suspected abnormal region, extracting multiple topographic layers of the region from the digital radiograph; determining features of the region in each of the layers; and applying inter-layer multivariate non-linear criteria to the features to determine whether the suspected abnormal region is to be classified as an abnormal region. The distribution of digital values in the digitized radiograph is modified and a probabalistic determination is made as to whether suspected abnormal regions are actually abnormal from multiple analyses. An image is processed using at least two partially correlated detection schemes to produce at least two corresponding result sets of potentially abnormal regions; and the results of the detection schemes are combined using logical 'AND' or logical 'OR' combining rules to detect abnormal regions in the radiograph.

Method and system to enhance robust identification of abnormal regions

Training/optimization of computer aided detection schemes based on

Method and apparatus for detecting abnormal regions such as masses and microcalcifications in living tissue depicted in radiographs such as mammograms is presented by digitizing the radiograph to produce a digitized radiograph, and, by computer, identifying, according to predetermined criteria, any suspected abnormal regions; and for each identified suspected abnormal region, extracting multiple topographic layers of the region from the digitized radiograph; and evaluating the extracted multiple topographic layers' abnormal regions to determine whether the region is an abnormal region; and displaying the abnormal regions in a version of the radiograph. Layers are extracted by thresholding the digital radiograph at a particular threshold value for each topographic layer. The threshold value for each other layer is based on the threshold and contrast values in a previous layer.

Computerized detection of masses and microcalcifications in digital

A method of detecting and identifying abnormal regions in living tissue depicted in a digital radiograph includes, in each of a number of stages, identifying an abnormal region; and reducing any overlap between the identified regions. Typically the digital radiograph is a mammogram and the abnormal regions are suspicious mass regions in a human breast. Five stages of identifying are used to detect a region with a global minimum in the smoothed image; a region with a local minimum in the original image; a region with a local minimum in the filtered image; a small region of rounded shape and low contrast; and a region of rounded shape and high contrast.

Growing community of inventors

Pittsburgh, PA, United States of America

Bin Zheng