This study plans to construct a MR radiomics model for predicting pathological complete response(pCR) to neoadjuvant chemoradiotherapy(CRT) in locally advanced rectal cancer(LARC) patients.

We propose to develop and validate a radiomics model for individualized pCR evaluation after CRT in patients with LARC. We plan to use both pre- and post-CRT MRI data to construct the predictive radiomics model for evaluating if LARC patients achieve pCR after CRT. The ultimate aim is to select appropriate LARC patients for omission of surgery.

Inclusion Criteria:

• pathologically proved rectal cancer
• locally advanced rectal cancer (≥T3 or N+)
• a distance less than 12cm between the lower edge of tumor and the anal margin
• no evidence of distant metastases
• no prior anti-cancer therapy before treatment
• scheduled to receive preoperative CRT

Exclusion Criteria:

• history or concurrent of other malignancy
• incomplete preoperative CRT
• failed to receive surgery or unavailable pCR assessment
• poor quality of MR images for measurement
• patient quit

LARC patients who will undergo neoadiuvant CRT before surgery.