Predicting the Risk of Breast Cancer using a Deep Learning Approach

Authors

Rishabh Shrivastava, Prof. Sudha Sharma

Abstract

After lung cancer, breast cancer is the second most common cancer. The most common cancer worldwide is lung cancer. On average, reproductive-age women are diagnosed with breast cancer more often than men. To minimise breast cancer fatalities, early identification is essential. Breast cancer causes are unknown, thus this is why. Early cancer detection may boost survival by 8%. This includes X-rays, mammograms, and MRI scans. How're things? Even the best doctors have trouble finding microscopic lumps, bumps, and masses, resulting in many false positives and negatives. This indicates nothing favourable. However, many seek to build better apps that may identify breast cancer early. This new technology can analyse photos and learn from them. A Deep Convolutional Neural Network (CNN) was utilised to distinguish carcinomas, calcifications, benign tumours, and abnormalities. Before, simple methods achieved this goal. To help doctors create better cancer treatments, the condition was classed as normal or aggressive. The model was trained before this. We started using this method to accomplish transfer learning efficiently. This is ResNet50. We upgraded our deep learning model similarly to our model. The learning rate of a neural network is crucial to training it. Use our strategy to adapt your study speed to match your needs. In the beginning of learning, mistakes are inevitable.

Keywords

Breast Cancer, CNN, Mammograms-MINI-DDSM, Machine Learning

Predicting the Risk of Breast Cancer using a Deep Learning Approach. Rishabh Shrivastava, Prof. Sudha Sharma. 2024. IJIRCT, Volume 10, Issue 2. Pages 1-11. https://www.ijirct.org/viewPaper.php?paperId=2404027