BREAST CANCER PREDICTION USING DEEP LEARNING TECHNIQUE

Abstract

Breast cancer is the second most prevalent kind of cancer globally, behind lung cancer. Breast cancer is the most common disease affecting women worldwide. There is always the potential for progress and development in the field of medical imaging. Cancer mortality is expected to reduce if the disease is detected and treated early and adequately. Doctors may be able to improve their diagnostic accuracy with the assistance of machine education. Deep learning, also known as neural networking, is one technique for distinguishing between benign and cancerous breasts. As a result, CNN may be used. The Mammograms-MINIDDSM data-collection used in this study consisted of 5358 mammograms, with about 2180 benign and 2998 malignant breast images being obtained. Deep learning in the diagnosis of mammogram cancer has been demonstrated in mammograms in promising experimental findings that encourage the use of in-depth learning techniques based on current characteristics and classification of a range of applications, particularly in the detection of breast cancer demonstrated in mammograms. Though work has to be done, improvements in CNN design and the use of pretrained networks allow for more development, which should result in improved accuracy over the next several years. To extract and categorize features efficiently, proper segmentation is required.