Related Experiment Video
Updated: Jun 26, 2025
10:41
Robotic D3 Partial Duodenal Resection with Primary Side-to-Side Anastomosis
Published on: December 15, 2023
1.8K
Artificial intelligence-powered clinical decision making within gastrointestinal surgery: A systematic review
Mustafa Bektaş1, Cevin Tan1, George L Burchell2
1Amsterdam UMC Location Vrije Universiteit Amsterdam, Surgery, De Boelelaan 1117, Amsterdam, the Netherlands.
Summary
Artificial intelligence (AI) can help optimize treatments in gastrointestinal surgery by predicting outcomes. Further prospective studies are needed to confirm the definitive role of AI in clinical decision-making for these complex cases.
Area of Science:
- Gastrointestinal Surgery
- Medical Artificial Intelligence
- Clinical Decision Support
Background:
- Clinical decision-making in gastrointestinal surgery is complicated by unpredictable tumor behavior and postoperative complications.
- Artificial intelligence (AI) offers potential to improve clinical decision-making by predicting surgical outcomes.
- The current literature on AI for clinical decision-making in gastrointestinal surgery requires synthesis.
Purpose of the Study:
- To provide an overview of AI models utilized for clinical decision-making in gastrointestinal surgery.
- To identify the applications and methodologies of AI in predicting surgical outcomes.
- To assess the current landscape of AI in gastrointestinal surgical practice.
Main Methods:
- Systematic literature search conducted across PubMed, EMBASE, Cochrane, and Web of Science.
- Inclusion criteria focused on studies using AI for clinical decision-making in gastrointestinal surgery patients.
- Exclusion criteria included reviews, pediatric studies, and study abstracts; methodological quality assessed using the Probast risk of bias tool.
Main Results:
- Ten articles were eligible from 1073 studies, highlighting AI applications in surgical procedure selection, chemotherapy, postoperative follow-up, and ileostomy implementation.
- Random Forest and Gradient Boosting models were most common, achieving Area Under the Curve (AUC) values up to 0.97.
- All included studies employed a retrospective design, with only one study performing external validation.
Conclusions:
- AI models demonstrate potential in selecting optimal treatments for gastrointestinal surgery patients.
- Integration of AI into clinical decision-making could yield significant clinical benefits.
- Prospective studies and randomized controlled trials are essential to establish the definitive role of AI in gastrointestinal surgery decision-making.
