Purpose/Significance The paper retrospectively analyzes the research progress on risk prediction models for prostate cancer, and provides references for the construction of prostate cancer intelligent diagnosis and treatment system.Method/Process Through literature mining and analysis, the significance of molecular, imaging, individual, population and other omics level evaluation indexes in the diagnosis of prostate cancer is discussed, and the key characteristic variables and clinical application value of different calculation models are compared.Result/Conclusion The existing models have advantages of easy calculation and strong feasibility, but they also have limitations such as limited prediction accuracy and insufficient generalization ability. The integration of multi-omics data and artificial intelligence models is of great significance for medical informatics research and smart medical system construction of prostate cancer.