Recognition and Segmentation of Gastric Tumor Based on Deep Residual Network
Haixia Zhao · 2020
In order to better complete the task of tumor segmentation, enterprises need to identify tumors in cardiac CT images. First, the CT image is preprocessed, including denoising, segmentation, amplification and gray value. In this paper, the deep residual network framework is built, and the pre-processed CT images are made into data sets and trained. The deep residual network is characterized by fast speed, good convergence and high robustness. The trained model is applied to the test data, and the predicted result and the weight coefficient are weighted as the final result, which is then used as the parameter to judge whether there is a tumor, so as to complete the tumor recognition.