Biliary Disease

Endoscopic retrograde cholangiopancreatography (ERCP (, a key tool that is used in diagnosis and treatment of pancreato-biliary diseases. Post-ERCP pancreatitis (PEP) is the most common and serious complication that can occur following this procedure and can lead to significant morbidity and mortality. A variety of patient-related and procedure-related factors have been associated with higher rates of PEP.

The rapid advancement of artificial intelligence (AI) has expanded its applications in healthcare, particularly in diagnostic assistance, intelligent triage, and patient interaction. Hepatobiliary and pancreatic diseases (such as liver cancer, pancreatic cancer, cirrhosis) are characterized by insidious onset, rapid progression, low early-diagnosis rates, and poor prognosis. However, grassroots medical institutions in China face challenges including physician shortages, variable patient health literacy, and incomplete initial information collection, leading to high misdiagnosis/missed diagnosis risks. Recent breakthroughs in large language models (LLMs) and multi-agent systems (MAS) offer new solutions. LLMs enable advanced natural language processing, while MAS coordinates specialized agents for complex decision-making. Integrating MAS with medical LLMs could create intelligent pre-consultation systems that systematically collect patient symptoms, risk factors, family history, and lifestyle data to enhance diagnostic efficiency. This study aims to develop a MAS-based pre-consultation system for hepatobiliary-pancreatic diseases featuring four specialized agents ("guidance agent," "medical history agent," "risk assessment agent," and "summary generation agent"). The system will simulate clinical reasoning to generate structured diagnostic reports for physicians. Research Objectives: Develop a specialized multi-agent framework combining LLMs to simulate clinical diagnostic logic and standardize symptom collection Enhance pre-consultation data integrity through intelligent dialogue focusing on key disease indicators Generate structured diagnostic summaries highlighting critical symptoms and risk factors Establish foundation for clinical validation and application through expert evaluation and user feedback This pre-diagnostic tool will assist physicians rather than replace clinical judgment, promoting safe, effective AI applications in early disease screening and tiered healthcare systems.

Choledocholithiasis (stone(s) in the common bile duct) is common. Untreated or missed, choledocholithiasis has high morbidity and mortality. Endoscopic retrograde cholangio-pancreatography (ERCP) is recognized as the first-line modality for management. While effective, ERCP is associated with adverse events. Thus, the selection of patients for ERCP should be accompanied by a high pre-test suspicion of choledocholithiasis. Choledocholithiasis is suspected based on clinical, biochemical and radiographic findings. The most relied-upon strategy for risk stratification of choledocholithiasis is based on guidelines from The American Society for Gastrointestinal Endoscopy (ASGE). In it, clinical predictors are defined as "very strong", "strong" or "moderate", and the presence of one or more of these is meant to suggest "high" or "intermediate" probability of choledocholithiasis. A knowledge gap exists in the performance characteristics of intermediate-probability criteria, where overall accuracy is \<50% from limited data. Patients in this group are recommended to a) undergo endoscopic ultrasound (EUS) or magnetic resonance cholangio-pancreatography (MRCP), b) undergo cholecystectomy with intra-operative cholangiography (IOC), or c) proceed directly to ERCP. At centres where EUS and MRCP are readily available, these are preferred options, as they are least invasive and sensitive; however, they are often unavailable. Thus, in clinical practice, a high proportion of intermediate-risk patients ultimately proceed directly to ERCP, where likelihood of benefit is only moderate, while procedural risk remains. The role of liver enzyme changes has not been evaluated; however, dynamic changes may offer another method for evaluating patients at intermediate risk of CBD stones that is safe and available. Incorporation of dynamic liver enzymes may improve the test-performance characteristics of the existing framework.

The aim of the Calgary Registry for Advanced and Therapeutic Endoscopy (CReATE) is to be a high-fidelity prospective multi-centre registry. The study population consists of consecutive adult ERCP patients from September 2018 to August 2022. Informed consent is acquired for each patient. All relevant pre-procedural, procedural, peri-procedural and post-procedural data are captured in real time by a full-time third-party research assistant directly observing procedures. Outcomes are ascertained by comprehensive medical record review and patient phone interview 30 days after the index procedure. This registry also serves as a secure data collection platform for several currently recruiting prospective studies and randomized trials.