AI models effectively predict successful sphincter preservation in low rectal cancer

A study involving 604 low rectal cancer patients revealed a 71.9% successful sphincter-preserving resection (SSPR) rate, with hospital rates ranging from 37.7% to 94.4%. Researchers developed seven AI algorithms, including logistic regression and light gradient boosting, achieving excellent discrimination for predicting SSPR, with Area Under the Receiver Operating Characteristic (AUROC) values exceeding 0.8 for the top models. An online calculator based on these findings was created to enhance clinical utility and decision-making.

• Why it matters: Variability in sphincter preservation rates needs efficient prediction solutions.

Randomized Controlled Trial by Wang X, Jiang W (…) Chi P et 9 al. in Int J Surg

Copyright © 2024 The Author(s). Published by Wolters Kluwer Health, Inc.

read the whole article in Int J Surg

open it in PubMed