Artificial intelligence and machine learning technologies in ulcerative colitis.

Area of Science:

• Gastroenterology
• Medical Informatics
• Computational Biology

Background:

• Growing interest in artificial intelligence (AI) for ulcerative colitis (UC) management.
• Over 80 studies in the last 5 years exploring machine learning (ML) for UC clinical problems.

Purpose of the Study:

• To review AI and ML applications in ulcerative colitis (UC).
• To highlight the use of AI/ML in UC diagnosis, prognosis, biomarker identification, disease monitoring, and complication prediction.

Main Methods:

• Utilized AI classifiers (e.g., random forest, support vector machines, neural networks) on molecular and clinical data.
• Applied ML algorithms (e.g., computer vision, convolutional neural networks) to analyze imaging data (endoscopic, histologic, radiologic).

Main Results:

• AI/ML models demonstrated high performance (accuracy, sensitivity, specificity) in modeling UC outcomes.
• ML algorithms successfully analyzed complex imaging datasets for UC diagnosis and complication prediction.

Conclusions:

• AI and ML tools offer a promising approach to optimize ulcerative colitis (UC) clinical practice.
• Large, high-quality validation studies are essential to mitigate bias and assess cost-effectiveness for clinical integration.