Radiomics prediction of surgery in ulcerative colitis refractory to medical treatment

1

Ulcerative colitis (UC) affects 156–291 per 100,000 people annually, leading to chronic inflammation and potential surgery in 8–24% of cases.
2

Surgical decisions for UC patients are complex and depend on individual factors, including drug history and patient preference.
3

Radiomics utilizes high-dimensional image analysis to predict treatment outcomes, showing promise in differentiating UC from Crohn's disease.
4

This study aims to use machine learning on CT images to predict whether UC patients will require surgery or can be treated medically.
5

The study employs LASSO for feature selection and evaluates predictive models using ROC curves to assess their accuracy in predicting surgery.

Techniques In Coloproctology

by K. Sakamoto, K. Okabayashi, R. Seishima, K. Shigeta, H. Kiyohara, Y. Mikami, T. Kanai, Y. Kitagawa
May 10, 2025

1