Using Machine Learning to Lower the Need for Diagnostic EUS or MRCP in Adults with Intermediate Risk of Choledocholithiasis

What this Trial Is About

Researchers are studying patients with suspected choledocholithiasis, a condition involving gallstones in the bile duct, who fall into an intermediate likelihood group based on current risk criteria. The study aims to evaluate how a machine learning model can help predict choledocholithiasis and potentially reduce the need for additional diagnostic procedures like Endoscopic Ultrasound EUS or Magnetic Resonance Cholangiopancreatography MRCP. This approach may lower healthcare use and costs for these patients. This observational study uses a machine learning-based predictive model to stratify patients who otherwise might undergo EUS or MRCP. Participants are those aged 18 to 80 years with intermediate risk of choledocholithiasis. The models performance in predicting the condition will be assessed, focusing on accuracy and receiver operating characteristic curve analysis within one month. Participants will be monitored through diagnostic evaluations including EUS or MRCP as needed, with data collected to validate the machine learning model. Researchers will measure the models accuracy and compare it to the diagnostic tests. The study will track outcomes over one month, assessing how well the model predicts the presence of choledocholithiasis and its potential to reduce unnecessary diagnostic procedures.

Application of Machine Learning Models to Reduce Need for Diagnostic EUS or MRCP in Patients With Intermediate Likelihood of Choledocholithiasis