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

Application of Machine Learning Models to Reduce Need for Diagnostic EUS or MRCP in Patients With Intermediate Likelihood of Choledocholithiasis- A Prospective, Open Label, Diagnostic Study

Status

Recruiting

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT06066372

Enrollment

1000

Registered

2023-10-04

Start date

2023-10-01

Completion date

2026-10-30

Last updated

2026-01-06

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Choledocholithiasis

Brief summary

Machine learning predictive model can help in stratifying heterogenous intermediate likelihood group to reduce need for EUS or MRCP in selected subgroup of patients.

Detailed description

The current guidelines for suspected choledocholithiasis are aimed to reduce the risk of patient receiving diagnostic ERCP and reduce the risk of post ERCP adverse events. In this process there is apparent increase in number of patients in the intermediate likelihood group requiring EUS or MRCP. This can increase the health care utilization and cost of care for intermediate likelihood patients. The field of artificial intelligence in clinical medicine is evolving rapidly. The use of artificial intelligence based machine learning model is not adequately studied for prediction of choledocholithiasis. Machine learning predictive model can help in stratifying heterogenous intermediate likelihood group to reduce need for EUS or MRCP in selected subgroup of patients.

Interventions

None listed

Study design

Observational model

CASE_ONLY

Time perspective

PROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to 80 Years

Inclusion criteria

• Individual 18 years or older with a suspected choledocholithiasis satisfying either ASGE or ESGE risk stratification criteria of intermediate likelihood undergoing EUS or MRCP

Exclusion criteria

* Patients having co-exiting disease of pancreato biliary system other than gall stones and choledocholithiasis which include chronic pancreatitis, biliary stricture, pancreatobiliary malignancy, portal biliopathy * Patients having underlying chronic liver diseases * Pregnancy and breast feeding * Previous history of cholecystectomy

Design outcomes

Primary

Measure	Time frame	Description
Area Under the Receiver Operating Characteristic Curve (AUROC) of the Machine Learning Model	1 month	Area under the receiver operating characteristic curve (AUROC) of the machine learning-based prediction model for identifying the presence of choledocholithiasis.

Secondary

Measure	Time frame	Description
Diagnostic Accuracy Metrics of Endoscopic Ultrasound (EUS) or Magnetic Resonance Cholangiopancreatography (MRCP)	1 Month	Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the receiver operating characteristic curve (AUROC) of magnetic resonance cholangiopancreatography (MRCP) for identification of choledocholithiasis.
Validation Performance of the Machine Learning Prediction Model	1 Month	Validation performance of the machine learning model for predicting choledocholithiasis, assessed using AUROC, calibration metrics (Brier score), and calibration plots in an independent validation cohort.

Countries

India

Contacts

Primary ContactNitin G Jagtap, MD

docnits13@gmail.com+919182859523

Backup ContactHardik Rughwani, MD

hardik.hr@gmail.com+919182859523

Outcome results