CT Image-based Breast Cancer Risk Detection Leveraging ResNet
Yifan An · 2023
Breast epithelial tissue can become a malignant tumour known as breast cancer. The second most lethal illness in women, it is a malignancy that affects many women. The breast cancer is now the number one malignant tumour in women, with an incidence rate of 100,000 per 100,000 in developed countries in Europe and the USA. It seriously affects women’s physical and mental health. Early detection, treatment and diagnosis are crucial for patients since they can increase their survival chances. In order to achieve recognition of breast cancer images, a neural network-based approach was chosen to investigate breast cancer images. Also compared with the traditional machine learning models. The image pre-processing process cropped a large number of blank areas and data enhancement was performed on positive data due to uneven data distribution. Finally, it was shown that ResNet can recognize breast cancer images well and the experimental results improved significantly (from 70.3% to 79.4%) after image processing and data enhancement.