直肠癌癌结节影像学评估现状与挑战:从解剖形态到人工智能OA

Tumor deposits(TDs)in rectal cancer represent a high-risk pathological feature independent of lymph node metastasis.They hold significant value in risk stratification and treatment decision-making for rectal cancer and have been incorporated into magnetic resonance imaging-based rectal cancer assessment criteria.Furthermore,quantitative func-tional imaging and integrated artificial intelligence(AI)models show potential in enhancing the diagnostic efficacy for TDs.The imaging evaluation system for rectal cancer is evolving from purely morphological descriptions toward a compre-hensive model that encompasses functional imaging,AI,and predictions of biological behavior.This article focuses on ana-lyzing the barriers to the clinical translation of advanced imaging technologies,the impact of deviations from the patho-logical"gold standard"on AI models,and the generalizability limitations of single-center data,aiming to provide an ob-jective reference for clinical diagnosis and treatment.