Breast Cancer Detection Based on Image Denoising in Multiple Modes
Zheng Yin, Shijie Pang, Yi Yang · Atlantis Highlights in Computer Sciences/Atlantis highlights in computer sciences · 2023
Breast cancer is cancer that develops from breast tissue, and it is the leading type of cancer in women.Convolutional neural network (CNN) is a very effective auxiliary method for medical image detection and classification as well as denoising, which is very important for diagnosis and analysis of medical images.In this study, DenseNet was used for breast cancer image classification and REDNet and a PRIDNet was sued for image denoising.By comparing the accuracy of different input with different noise level and denoising model, this study showed that denoising can remove the redundant information of images with noise and improve the accuracy of classification and higher Peak Signal to Noise Ratio (PSNR) and Structural Similarity (SSIM) would lead to a higher classification accuracy.The model performed better on the images with similar noise to the noise of the training image.