Owing to their insidious onset and atypical early symptoms, ureteral tumors are prone to missed diagnoses and misdiagnoses. Accurate diagnosis is crucial for improving patient prognosis. Multimodal imaging techniques provide comprehensive diagnostic evidence through their complementary advantages. Among them, computed tomography urography (CTU) has become the core method for lesion detection and staging due to its high spatial resolution; multi-parametric magnetic resonance urography (MRU) has unique advantages in soft tissue evaluation; ultrasonography and positron emission tomography-computed tomography (PET/CT) play important roles in preliminary screening and metastatic lesion detection, respectively. AI-based radiomics enhances diagnostic accuracy and efficiency by extracting quantitative features from images and integrating machine learning algorithms to facilitate the differentiation of benign and malignant tumors, tumor grading and staging, as well as prognostic evaluation. This review summarizes the application status of multimodal imaging techniques and the technical progress of AI-based radiomics, analyzes the clinical value and existing challenges of their combined application, and provides a reference for the accurate diagnosis of ureteral tumors.