Improvement of Image Enhancement Technique for Mammography Images

Abstract

Algorithms of medical image analysis are important in the diagnosis and detection of breast cancer at an early stage. In this research work, the issue of breast cancer detection and diagnosis using topographic map-based segmentation and machine learning methods in mammography images is considered. Image preprocessing and segmentation algorithms have been improved to solve this problem. Based on improved algorithms, classification results were obtained with SVM, Random Forest, ANN, k-NN, and Decision tree classifiers. The MIAS image dataset was used to analyze the efficiency of the developed algorithms.