Siemens Medical Solutions Usa, Inc.

Murat Dundar

R Bharat Rao

Glenn Fung

Balaji Krishnapuram

Jinbo Bi

Sriram Krishnan

Vikas C Raykar

A method for training a classifier for classifying candidate regions in computer aided diagnosis of digital medical images includes providing a training set of images, each image including one or more candidate regions that have been identified as suspicious by a computer aided diagnosis system. Each image has been manually annotated to identify malignant regions. Multiple instance learning is applied to train a classifier to classify suspicious regions in a new image as malignant or benign by identifying those candidate regions that overlap a same identified malignant region, grouping each candidate region that overlaps the same identified malignant region into a same bag, and maximizing a probability wherein N is a number of bags, pis a probability of bag i containing a candidate region that overlaps with an identified malignant region, and yis a label where a value of 1 indicates malignancy and 0 otherwise.

System and method for multiple-instance learning for computer aided diagnosis

A method of training a classifier for computer aided detection of digitized medical image, includes providing a plurality of bags, each bag containing a plurality of feature samples of a single region-of-interest in a medical image, where each region-of-interest has been labeled as either malignant or healthy. The training uses candidates that are spatially adjacent to each other, modeled by a 'bag', rather than each candidate by itself. A classifier is trained on the plurality of bags of feature samples, subject to the constraint that at least one point in a convex hull of each bag, corresponding to a feature sample, is correctly classified according to the label of the associated region-of-interest, rather than a large set of discrete constraints where at least one instance in each bag has to be correctly classified.

System and method for multiple instance learning for computer aided detection

A method for training classifiers for Computer-Aided Detection in medical images includes providing an image feature training set {(x, y)}, wherein x&#x3b5;Rare input feature variables and y&#x3b5;{&#x2212;1,1} are class labels, and a cascade of K classifiers to be trained, minimizing, for each classifier k, a first cost function to initialize an &#x3b1;associated with each classifier k, fixing all classifiers except classifier k and minimizing a second cost function to solve for &#x3b1;for a counter value c using the training dataset {(x, y)}, calculating a third cost function J(&#x3b1;, . . . , &#x3b1;) for each classifier k, and comparing Jwith a previous iteration J, wherein if J&#x2212;Jis less than a predetermined tolerance, said classifier training is completed.

System and method for joint optimization of cascaded classifiers for computer aided detection

CAD (computer-aided diagnosis) systems and applications for breast imaging are provided, which implement methods to automatically extract and analyze features from a collection of patient information (including image data and/or non-image data) of a subject patient, to provide decision support for various aspects of physician workflow including, for example, automated diagnosis of breast cancer other automated decision support functions that enable decision support for, e.g., screening and staging for breast cancer. The CAD systems implement machine-learning techniques that use a set of training data obtained (learned) from a database of labeled patient cases in one or more relevant clinical domains and/or expert interpretations of such data to enable the CAD systems to 'learn' to analyze patient data and make proper diagnostic assessments and decisions for assisting physician workflow.

Systems and methods for automated diagnosis and decision support for breast imaging

A method and device having instructions for analyzing input data-space by learning classifiers include choosing a candidate subset from a predetermined training data-set that is used to analyze the input data-space. Candidates are temporarily added from the candidate subset to an expansion set to generate a new kernel space for the input data-space by predetermined repeated evaluations of leave-one-out errors for the candidates added to the expansion set. This is followed by removing the candidates temporarily added to the expansion set after the leave-one-out error evaluations are performed, and selecting the candidates to be permanently added to the expansion set based on the leave-one-out errors of the candidates temporarily added to the expansion set to determine the one or more classifiers.

Growing community of inventors

Malvern, PA, United States of America

Murat Dundar