Novel Fuzzy Topological Approach to the Detection of Mammographic Lesions and Quantification of Parenchymal Density

摘要

This research focuses on mammographic image processing for the purpose of density quantification, lesion detection and quantification. The approaches are different from those taken in the literature in two respects. (1) They emphasize on identifying the dense regions and analyzing the parenchymal architecture. (2) They use a novel fuzzy connected-ness method of object definition and image segmentation. During this project period, the following have been accomplished: The development and validation of a new method of lesion and density detection based on fuzzy connectedness that utilizes the relative strength of connectedness among objects. The development of a new class of interactive segmentation methods that are more efficient and effective for lesion segmentation than the earlier live wire methods. Further extension of the fuzzy connectedness framework and demonstration of its effectiveness in areas other than mammographic image processing.