Artificial Intelligence Helps Predict Ulcerative Colitis Flare-ups, Prognosis


Authors of a study published in Gastroenterology say results show that an artificial intelligence-based, computer-aided model accurately differentiates disease remission from inflammation.

Ulcerative colitis is a chronic inflammatory bowel disease (IBD) that follows a relapsing-remitting pattern that can be quite unpredictable. Treatment focuses on eliminating inflammation and promoting histologic remission. Histopathology is considered the most accurate method of detecting inflammation and differentiating it from remission.

Artificial Intelligence (AI)-based, computer-aided diagnosis (CAD) systems are gaining popularity to simplify and standardize the assessment of medical imaging.

Marietta Iacucci, M.D., Ph.D., led a study of ulcerative colitis that combined artificial intelligence and computer-aided diagnosis.

Marietta Iacucci, M.D., Ph.D., led a study of ulcerative colitis that combined artificial intelligence and computer-aided diagnosis.

In a study published earlier this month in Gastroenterology, Marietta Iacucci, M.D., Ph.D., from Birmingham University Hospitals, United Kingdom, and colleagues sought to develop and validate an AI-based CAD system to assess ulcerative colitis biopsy samples and predict disease prognosis.

Iacucci and her colleagues recruited patients from 11 international centers between September 2016 and November 2019. Eligible participants had a confirmed diagnosis of ulcerative colitis for at least one year without regard to disease activity and an indication for a colonoscopy. At least two tissue samples were obtained from the rectum and the sigmoid because they are common areas representative of healing and inflammation. The endoscopic exam was recorded in the same area.

Clinical outcomes used as proxies for disease flare-ups for the purpose of prognosis assessment included ulcerative colitis-related hospitalizations or surgery and increase in initiation of or changes in ulcerative colitis treatments, such as immunomodulators, biologics, or steroids, due to worsening symptoms. These outcomes were recorded during follow-up phone calls or visits 12 months after the endoscopy in the initial group or up to 33 months for the external validation group.

A total of 535 samples from 273 participants were used to develop and test the model. Of these, 118 biopsy samples were used to train the model, 42 were used to calculate it, and 375 to test it. The CAD system was trained to detect neutrophils and predict the PICaSSO Histologic Remission Index (PHRI). PHRI is an index used to gauge clinical outcomes based on endoscopic activity in ulcerative colitis.

The researchers found that the CAD system differentiated histologic remission from disease activity, as detailed by PHRI, with 89% sensitivity, 85% specificity, 75% positive predictive value, 94% negative predictive value, and 87% accuracy. When compared with human evaluations, the AI-based CAD system showed comparable results.

Iacucci and colleagues conclude that their AI-based CAD model accurately distinguishes disease remission from inflammation and provides a good tool for predicting the risk of flare-ups. They wrote, “Our computer tool can speed up, simplify, and standardize histological assessment of ulcerative colitis in clinical practice and clinical trials, and provide accurate prognostic information to the clinician.”

Related Content
© 2024 MJH Life Sciences

All rights reserved.