This study employed MR cytometry imaging combined with deep-radiomic research methods to predict the heterogeneity of rectal cancer, and constructed a model to predict the sensitivity to neoadjuvant therapy and potential mechanisms, assisting clinicians in early identification of high-risk tumors and patients with treatment-susceptible rectal cancer, and selecting individualized treatment plans, thereby improving patient prognosis. To this end, the investigators need to collect the clinical diagnosis and treatment data of patients, surgical and biopsy pathological information, pre-treatment rectal MRI images and colonoscopy biopsy tissue sections, and follow up on the surgical pathological results and recurrence of patients. This study is a non-interventional diagnostic study, which does not interfere with the routine diagnosis and treatment of patients, does not affect any medical rights of patients, and does not increase any medical risks.

Inclusion Criteria:

• Newly diagnosed rectal adenocarcinoma without previous treatment
• Complete radical surgery in the institute
• Complete rectal MRI within 2 weeks before surgery

Exclusion Criteria:

• Concomitant malignancies or systemic disease
• Inadequate MR image quality for analysis
• Incomplete clinicopathological data
• Loss of follow-up

Chinese patients diagnosed with rectal cancer through pathological examination