ENHANCING EARLY DETECTION OF BREAST CANCER WITH PRE-PROCESSING TECHNIQUES

Abstract

                          Mammogram picture segmentation is crucial for early identification of breast cancer. However, because of the intrinsic complexity and diversity of mammography images, reliable segmentation remains difficult. Noise frequently degrades mammogram pictures, making precise diagnosis and segmentation of breast cancer difficult. In this research, we present a novel technique to breast cancer diagnosis that comprises of two stages: Hybrid CLEACH with wavelet denoising for noise reduction and genetic algorithms-based Contourlet Transform for picture segmentation. In the first stage, wavelet denoising is used to reduce noise from mammography pictures. In the second stage, image segmentation is performed using genetic algorithms based on the Contourlet Transform to accurately detect and segment breast cancer. The Contourlet Transform performs a multiresolution examination of the picture, improving the edges and details of the mammography, while the genetic algorithm optimizes the segmentation parameters. The suggested method is assessed using mammography images from the MIAS datasets and outperforms state-of-the-art methods in terms of segmentation accuracy, sensitivity, and specificity. The proposed strategy could be employed in clinical practice to increase breast cancer detection accuracy while reducing false positives and false negatives.