System and method for automated segmentation, characterization, and classification of possibly malignant lesions and stratification of malignant tumors

摘要

A method and apparatus for classifying possibly malignant lesions from sets of DCE-MRI images includes receiving a set of MRI slice images obtained at respectively different times, where each slice image includes voxels representative of at least one region of interest (ROI). The images are processed to determine the boundaries of the ROIs and the voxels within the identified boundaries in corresponding regions of the images from each time period are processed to extract kinetic texture features. The kinetic texture features are then used in a classification process which classifies the ROIs as malignant or benign. The malignant lesions are further classified to separate TN lesions from non-TN lesions.

主权项

1. A method for classifying a possibly malignant or non-malignant region of interest (ROI) from sets of dynamic contrast-enhanced (DCE) MRI images acquired over a series of time periods, the method comprising:
receiving a plurality of ROI slices, each ROI slice representing a respective one of the series of time periods, and corresponding to one of the sets of DCE MRI images, each of the ROI slices including a plurality of voxels representative of the ROI in the corresponding time period; automatically extracting a boundary of the ROI for each received ROI slice, by processing the voxels in the received ROI slice according to an expectation/maximization algorithm to group the voxels and by processing the grouped voxels using a magnetostatic active contour model to extract the boundary of the ROI; for each ROI slice, determining texture features of the voxels within the boundary of the ROI, to the exclusion of the voxels outside of the boundary; determining kinetic texture features of the ROI based on the respective extracted boundary of the ROI and the texture features in each ROI slice, the kinetic texture features being representative of spatio-temporal changes in the texture features of each of the voxels in the ROI over the series of time periods; and applying a classifier to the kinetic texture features to classify the ROI as a whole as belonging to one of at least two classes.