Research on the Prediction of Benign and Malignant Ovarian Tumors Using Improved ResNet
Yuhao Lang, Qiudong Yu, Xiaohang Sun, Yuhan Cheng · 2024
With the wide application of deep learning in medical image analysis, the prediction of benign and malignant ovarian tumors has become an important research direction in the medical field. This study improves ResNet (Residual Networks) to improve its accuracy in the prediction task of benign and malignant ovarian tumors. Firstly, we introduce parallel downsampling and CBAM hybrid attention mechanism, which enables the model to capture a wider range of contextual information and enhances the model's attention to important features, improving the ability to extract information from key areas of ovarian tumors. Secondly, through the improved activation function layer, we optimize the efficiency of feature information transmission, making the model better adapt to the nonlinear characteristics of ovarian tumor images. In addition, we adopt an appropriate data augmentation strategy to expand the training dataset, improve the generalization performance of the model, and provide more reliable support for the accurate diagnosis of ovarian tumors.