Comparative Analysis of Different Preprocessing Techniques for Digital Mammograms Towards Breast Cancer Detection

Breast cancer, finds itself as the most dreaded and commonly found cancer worldwide. The CAD system using mammograms assists radiologists for breast tumor diagnosis in its earlier stages. Preprocessing mammographic images is first action in tumor identification, as it contributes to image quality enhancement, noise reduction, feature extraction, and artifact removal. Precise segmentation of breast and pectoral muscle regions is another step-in preprocessing as the pectoral muscle appears the same bright as cancerous tissue on a mammograph, giving a false positive indication and changing the direction of the medication. The paper’s Prime objective is to offer a comparative study of different pre-preprocessing and pectoral muscle separation methods, for breast cancer detection. During this review, we encountered different noise removal, image enhancement and pectoral muscle removal methods. Adaptive median filtering is most popular method is utilized in removal of noise. Even though various approaches are implemented for the preprocessing of digital mammogram accuracy, precision is not reached at its optimal level. This review article provides merits, demerits and future scope of different preprocessing and segmentation techniques, which will be helpful for readers and future researchers to design and develop novel preprocessing and segmentation techniques which will be beneficial in the accurate detection of cancer.