Lung cancer, breast cancer, liver cancer, and glioma are prevalent malignant tumors. Leveraging collected multimodal clinical big data, our team has developed diagnostic models such as automated lung nodule detection and Lung-RADS classification to achieve early and precise diagnosis of these common tumors. For treatment efficacy evaluation, we proposed a glioma segmentation algorithm based on multimodal feature fusion. Furthermore, we have successfully developed software for constructing multimodal databases for common tumors and an AI-PACS system, which are currently being implemented and promoted in over 20 hospitals, including the First Affiliated Hospital of Zhengzhou University.