Prostate cancer is one of the most common malignant tumors of the genitourinary system in elderly men.Accurate preoperative tumor grading can further improve the diagnosis and treatment outcomes of prostate cancer patients.Magnetic resonance image(MRI), as the main tool for preoperative diagnosis of prostate cancer, has shown an important value for the diagnosis and treatment of prostate cancer in clinical practice.However, due to differences in imaging equipment, examination methods and expertise in diagnosis among different medical institutions in China, the potential value of MRI in the diagnosis and treatment of prostate cancer has not been fully realized.Therefore, more advanced quantitative analysis methods are urgently needed to further expand the contribution of MRI in precise diagnosis and treatment of prostate cancer.The application of artificial intelligence(AI) technology in the analysis of magnetic resonance imaging of prostate cancer can help improve the quality and evaluation capability of quantitative analysis, and has great potential in accurate grading of prostate cancer.However, image-based AI model building is usually complicated, especially in the stages of data set preparation, model construction and model evaluation, which still need to be integrated and standardized.This consensus aims to promote the standardization of the whole process of model construction for AI-assisted precise grading of prostate cancer based on magnetic resonance imaging, so as to improve clinical precision grading of prostate cancer.