Lesion detection from the mammography image using the Vision Development Module

Abstract

In this study, it was aimed to detect lesion and their radius information from mammography images using the Vision Development Module. The study was performed using ten different images with uniformly distributed lesions from the mammography images found in the MIAS database. First, threshold was applied to the mammography images loaded into the Vision Development Module to make the lesions clear. Then morphological filtering, histogram equalization and shape detection functions were used consequently radius information was obtained by detecting lesions. As a result, the error value between the known radius values of the lesions and the radius values found using the module was calculated and the minimum error value was found to be 3% and the maximum error value was found to be 24%. The total error value in the determination of the radius of the lesion in all mammography images examined, was found to be 15%.