Predicting the Instance of Breast Cancer within Patients using a Convolutional Neural Network
(1) Aspiring Scholars Directed Research Program
Breast cancer is a widespread disease that affects millions of people across the world. This makes early detection and diagnosis critical for heightened chance of survival. Although there are many ways for breast cancer to be detected, many patients are diagnosed late in their disease progression. In this paper, we present a robust and automated way of early breast cancer detection within patients using machine learning methods. Our method uses medical images of possible breast cancer and classifies them as either non-cancerous or cancerous using a convolutional neural network (CNN). Our results showed that our CNN model had a mean accuracy of 97.25% when averaged using the final 50 percent of samples. These results show that machine learning can be used as an effective way to clinically diagnose breast cancer due to the high accuracy.
This article has been tagged with: