Related Experiment Video
Updated: Jul 12, 2025
06:09
Author Spotlight: Advancing Hepatic Fibrosis Diagnosis Using Magnetic Resonance Elastography and AI
Published on: July 21, 2023
1.2K
Predictive Algorithm for Hepatic Steatosis Detection Using Elastography Data in the Veterans Affairs Electronic
Saroja Bangaru1,2, Ram Sundaresh3, Anna Lee3
1Vatche and Tamar Manoukian Division of Digestive Diseases, David Geffen School of Medicine at the University of California, Los Angeles, Los Angeles, CA, 90095, USA.
Digestive Diseases and Sciences
|October 21, 2023
Summary
A new algorithm predicts nonalcoholic fatty liver disease (NAFLD) using clinical factors and controlled attenuation parameter (CAP) scores. This tool helps identify at-risk Veterans for early intervention and specialized care.
Area of Science:
- Hepatology
- Medical Informatics
- Public Health
Background:
- Nonalcoholic fatty liver disease (NAFLD) is a growing global health concern.
- Early detection of hepatic steatosis (HS) is crucial for timely hepatology referral.
- Vibration-controlled transient elastography (VCTE) with controlled attenuation parameter (CAP) is a validated non-invasive tool for HS diagnosis.
Purpose of the Study:
- To develop and validate a novel clinical predictive algorithm for HS in Veterans.
- To identify key clinical predictors of elevated CAP scores.
- To establish an optimal CAP threshold for HS detection in the Veteran population.
Main Methods:
- Retrospective analysis of 403 Veterans with VCTE data from the Greater Los Angeles VA Healthcare System.
- Exclusion of patients with alcohol-associated liver disease, specific hepatitis C genotypes, malignancies, or liver transplantation.
- Linear regression to identify predictors of NAFLD; receiver operating characteristic analysis to determine CAP thresholds using biopsy, MRI, and ultrasound as gold standards.
Main Results:
- The cohort was diverse: 26% Black/African American, 20% Hispanic.
- Predictors of elevated CAP included diabetes, cholesterol, triglycerides, BMI, and Hispanic ethnicity.
- The predictive model showed strong correlation (r=0.61) with actual CAP scores, with 82% sensitivity and 83% specificity in validation. Optimal CAP cutoff was 273.5 dB/m (AUC=75.5%).
Conclusions:
- A novel clinical algorithm effectively predicts HS in Veterans, aiding in risk stratification.
- This tool facilitates linking at-risk Veterans to non-invasive testing and sub-specialty care.
- Future research should explore the algorithm's applicability in non-specialty clinics due to potential referral biases.
