Work

Preoperative magnetic resonance imaging–based radiomics nomogram model for predicting postoperative anastomotic leakage in colorectal cancer

Year: 2025

Type: article

Abstract: PURPOSE: Anastomotic leakage (AL) is a serious postoperative complication after colorectal cancer surgery, and accurate preoperative prediction remains challenging. This study aimed to develop and validate a magnetic resonance imaging (MRI)-based radiomics nomogram for the preoperative prediction of AL. METHODS: A total of 146 patients with colorectal cancer, including 11 with AL, were retrospectively enrolled and randomly divided into training and validation cohorts at a 7:3 ratio. Clinical variables and preoperative MRI-based radiomic features were analyzed. A clinical model was constructed using logistic regression. Radiomic features were selected using the least absolute shrinkage and selection operator method to develop a radiomics model, from which a radiomic score was calculated. A combined radiomics nomogram integrating the radiomic score and significant clinical factors was subsequently established. Model performance was evaluated using receiver operating characteristic curve analysis in both cohorts. RESULTS: The clinical model achieved an area under the curve (AUC) of 0.766 in the training cohort and 0.583 in the validation cohort. The radiomics model demonstrated improved discrimination, with AUCs of 0.822 and 0.800, respectively. The combined radiomics nomogram showed the best predictive performance, yielding AUCs of 0.869 in the training cohort and 0.858 in the validation cohort. CONCLUSION: The proposed MRI-based radiomics nomogram demonstrates good predictive performance for postoperative anastomotic leakage and may serve as a useful tool for preoperative risk stratification in patients with colorectal cancer.

Source: Annals of Coloproctology

DOI: 10.3393/ac.2025.00689.0098

Authors: Junfeng Yan, Qiushuang Wang, Qiang Li, Jiatong Lu, Qiang Tong

Institution: Renmin Hospital of Wuhan University

Language: English

Topic: Radiomics and Machine Learning in Medical Imaging

Subfield: Radiology, Nuclear Medicine and Imaging

Field: Medicine

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $500