Siemens Medical Solutions Usa, Inc.

Siemens Aktiengesellschaft

Lin Hong

Hong Shen

Shuping Qing

Ying Li

Christopher Vincent Alvino

Yumao Lu

Yonggang Shi

A method and system for nodule feature extract using background contextual information in chest x-ray images is disclosed. In order to detect false positives in nodule candidates for a chest x-ray image, background contextual information, such as contextual vessel tree information, is defined in the chest x-ray image. Features are extracted for each nodule candidate based on the background contextual information, and the extracted features are used to detect whether each nodule candidate is a false positive or a genuine nodule.

Method and system for nodule feature extraction using background contextual information in chest x-ray images

Feature processing is provided for lung nodules in computer-assisted diagnosis. A feature that may better distinguish nodules from background is extracted using a Hough transform. Rather than relying on a specific boundary shape, the Hough transform accumulates evidence associated with a region, such as a ring region. The accumulated evidence provides a feature score without requiring a nodule to fit a specific shape. In another approach, a background level is determined from extracted features. Rather than attempting to normalize an image prior to extraction, the features are normalized. The feature normalization and generalized Hough transform extraction may be used together or alone.

Feature processing for lung nodules in computer assisted diagnosis

A method and system for locating an opaque region, such as a heart region in a chest X-ray radiograph is disclosed. In order to segment a heart region in a chest X-ray radiograph, a heart region boundary is generated based on lung boundaries in the chest X-ray radiograph and an average heart region model. A location of the lower boundary of the heart region in the chest X-ray radiograph is then determined. Left and right portions of the heart region boundary are independently registered to corresponding portions of the lung boundaries, and upper and lower portions of the heart region boundary are adjusted based on the left and right portions in order to form a smooth contour.

Method and system for locating opaque regions in chest X-ray radiographs

A method for performing computer-assisted lesion assessment includes acquiring contrast enhanced MRI data. A region of interest is identified. Kinetic properties, morphological properties and pattern of enhancement are determined for the identified region of interest. A score is calculated for each of the kinetic properties, morphological properties, pattern of enhancement. A cumulative score is calculated for the identified region of interest. The scores are provided to a medical practitioner to assist in diagnosing the identified region of interest.

Computer aided lesion assessment in dynamic contrast enhanced breast MRI images

A method and system for segmenting diaphragm regions in a chest X-ray radiograph is disclosed. The diaphragm regions are segmented based on left and right lung lobe regions and a heart region in the chest X-ray radiograph. A right diaphragm region is segmented in the chest X-ray radiograph based a boundary of the right lung lobe. A left diaphragm region is segmented in the chest X-ray radiograph based on the heart region and a boundary of the left lung lobe.

Method and system for diaphragm segmentation in chest X-ray radiographs

A system for segmenting a target organ tumor from an image includes a background model builder, a foreground model builder and a tumor region locator. The background model builder uses an intensity distribution estimate of voxels in an organ region in an image to build a background model. The foreground model builder uses an intensity distribution estimate of voxels in a target organ tumor to build a first foreground model. The tumor region locator uses the background model and the first foreground model to segment the target organ tumor to obtain a first segmentation result.

Systems and methods for segmenting object of interest from medical image

Certain exemplary embodiments can comprise a method, which can comprise automatically rendering an improved image of a target object. The improved image obtained based upon a principal mode of the target object. The principal mode of the target object can be provided to an algorithm that is adapted to derive the improved image of the target object.

Devices, systems, and methods for improving image consistency

In a method of 3-D object detection, an image is pre-processed. Using cross-section analysis, a confidence array is built. A plurality of peaks in the confidence array are detected, wherein the peaks signify a likelihood of a 3-D object of interest.

Object detection using cross-section analysis

Detecting ground glass nodules in medical images includes calculating a probability distribution function of background lung tissue in a chest image, estimating a variation range of the background tissue probability distribution function, estimating a probability distribution function of an image point belonging to a ground glass nodule from the variation range of the background tissue probability distribution function by using a function corresponding to the variation range of the background tissue probability distribution function, and calculating a log likelihood function of the image from the background tissue probability distribution function and the estimated ground glass nodule probability distribution function, wherein the log likelihood function represents the confidence that a point in the image is not part of a ground glass nodule. The log likelihood function value for each point is compared to a confidence value of the background tissue, to determine if the point is a candidate ground glass nodule location.

System and method for detecting ground glass nodules in medical images

A computer-implemented method for candidate generation in three-dimensional volumetric data comprises forming a binary volumetric image of the three-dimensional volumetric data including labeled foreground voxels, estimating a plurality of shape features of the labeled foreground voxels in the binary volumetric data including, identifying peak voxels and high curvature voxels from the foreground voxels in the binary volumetric image, accumulating a plurality of confidence values for boundary and each peak voxel, and detecting confidence peaks from the plurality of confidence values, wherein the confidence peaks are determined to be the candidate points, and refining the candidate points given detected confidence peaks, wherein refined candidate points are determined to be candidates.

Growing community of inventors

Monmouth Junction, NJ, United States of America

Lin Hong