IFCM clustering segmentation based on genetic algorithm

Abstract

In the early diagnosis of breast cancer(BC), computer aided design (CAD) is particularly important, and the accurate detection of breast mass in mammography plays an important role Objective: In order to distinguish the mass region from other background regions, an effective segmentation scheme was proposed in Mammographic Image Analysis Society (MIAS) segmentation Methods: First, the initial clustering center of intuitionistic fuzzy C-means (IFCM) is determined by genetic algorithm (GA), and then the image is segmented by IFCM algorithm to turn the random initial clustering center into a purpose-selected one, so as to ensure the optimal result of the final clustering center results: The average segmentation precision of MIAS.I images with noise level of 5%,7% and 9% were 90.15% and 86.85% and 87.31%. Conclusion: This method combines the advantages of the two algorithms to segment the location of mass more accurately and quickly.