DEEP CONVOLUTIONAL NEURAL NETWORKS FOR AUTOMATED BREAST CANCER DIAGNOSIS

Abstract

Breast cancer is a disease where the cells in the breast grow out of control and form a tumor. It usually starts in the milk ducts or the glands that produce milk. If these abnormal cells are not treated, they can spread to parts of the body through the blood and lymph vessels, which can be very dangerous. This process is called metastasis. Breast cancer is more common in women. It is one of the leading causes of death among women in countries where they do not have healthcare. When diagnosis is delayed in these countries, it leads to outcomes. Finding breast cancer early is very important to improve the chances of survival in the fight against breast cancer. The tools we have now to diagnose breast cancer, like mammograms, have two problems:

Different doctors often do not agree on what the results mean.

They often give alarms, which causes patients a lot of stress and extra tests that they do not need.

This research is trying to find out how well a special computer system can identify and assess the risk of breast cancer from pictures. The computer system, called DCNN-BCD, uses learning to help medical professionals detect breast cancer early and correctly.

We tested the DCNN-BCD system on pictures to create an accurate tool to detect breast cancer automatically. The goal is to help doctors detect breast cancer correctly and early. Breast cancer detection is very important for improving survival rates. The computer system can help doctors with breast cancer diagnosis. That is very important for breast cancer. Breast cancer detection is crucial. The computer system can help doctors detect breast cancer.