Research
The Data Analytics and AI for Advanced Liver Disease Lab is pursuing several avenues of investigation. Highlights of the lab's research projects include developments in several areas.
ECG-based detection of advanced liver disease
The AI-Cirrhosis-ECG (ACE) score is a deep-learning model that can accurately detect the presence of cirrhosis and grade the severity of liver disease on digitized 12-lead ECGs. The ACE score has demonstrated an outstanding classification performance for distinguishing cirrhosis from no cirrhosis (area under the curve of 0.908). In addition, the magnitude of the ACE score closely associates with clinical disease severity and liver-related outcomes in cirrhosis.
The score was implemented to detect advanced liver disease at earlier stages than decompensated cirrhosis. It was subsequently evaluated in the Detection of Undiagnosed Liver Cirrhosis via Artificial Intelligence-Enabled Electrocardiogram (DULCE) pragmatic clinical trial. This clinical trial demonstrated that the ECG-AI model can effectively enhance the detection of liver disease in the general population. This includes both advanced stages and early fibrosis.
Publications
- Olofson A, Lennon R, Kassmeyer B, Liu K, Attia ZI, Rushlow D, Rattan P, Ahn JC, Friedman PA, Allen A, Kamath PS, Shah VH, Noseworthy PA, Simonetto DA. Detection of undiagnosed liver cirrhosis via artificial intelligence-enabled electrocardiogram (DULCE): Rationale and design of a pragmatic cluster randomized clinical trial. Contemporary Clinical Trials Communications. 2025.
- Rattan P, Ahn JC, Chara BS, Mullan AF, Liu K, Attia ZI, Friedman PA, Allen A, Shah VH, Kamath PS, Noseworthy PA, Simonetto DA. Training and performance of an electrocardiogram-enabled machine learning model for detection of advanced chronic liver disease. American Journal of Gastroenterology. 2025.
- Simonetto DA, Rushlow D, Liu K, Calleri A, Kassmeyer B, Lennon RJ, Rattan P, Bernard ME, Singh G, Deyo-Svendsen ME, King G, Stacey SK, Olofson A, Allen A, Ahn JC, Friedman PA, Kamath PS, Attia ZI, Noseworthy P, Shah VH. Detection of undiagnosed liver cirrhosis via AI-enabled electrocardiogram: A pragmatic, cluster-randomized clinical trial. Accepted, Nature Medicine. 2025.
- Ahn JC, Attia ZI, Rattan P, Mullan AF, Buryska S, Allen AM, Kamath PS, Friedman PA, Shah VH, Noseworthy PA, Simonetto DA. Development of the AI-Cirrhosis-ECG Score: An electrocardiogram-based deep learning model in cirrhosis. American Journal of Gastroenterology. 2022.
Digital phenotyping in alcohol-associated liver disease and alcohol use disorder
Improving our understanding in the area of digital phenotyping in alcohol-associated liver disease and alcohol use disorder can help us predict clinical outcomes, including cravings and relapse events.
This pilot study uses readily available technologies to identify signals of association between digital biomarkers and predictors of relapse. The study has demonstrated that digital phenotyping can detect meaningful signals correlating with relapse predictors, supporting its feasibility as a tool for continuous, real-time monitoring.
These findings highlight the potential of digital phenotyping to inform patient-centered, personalized management strategies for people with liver disease.
Publication
Remote patient monitoring for improved health outcomes in cirrhosis
This project is designed to reduce hospital readmissions among patients with decompensated cirrhosis. It involves at-home monitoring that includes daily vital signs and symptom questionnaires. The project is overseen by trained nursing staff who can implement targeted interventions such as medication adjustments or escalation to physician judgment.
In a pilot study, we demonstrated the feasibility of the remote patient monitoring program, showing that integrating digital health tools into routine hepatology care is both feasible and well accepted by patients. The remote patient monitoring intervention also was associated with meaningful reductions in hospitalizations and mortality, suggesting that early detection of clinical deterioration and timely intervention may alter the trajectory of advanced liver disease.
Publication
- Penrice DD, Hara KS, Sordi-Chara B, Kezer C, Schmidt K, Kassmeyer B, Lennon R, Rosedahl J, Roellinger D, Rattan P, Williams K, Kloft-Nelson S, Leuenberger A, Kamath PS, Shah VH, Simonetto DA. Design, implementation and impact of a cirrhosis-specific remote patient monitoring program. Hepatology Communications. 2024.
Application of virtual reality in patients with liver cirrhosis
The Virtual Reality study was a proof-of-concept pilot study designed to explore the feasibility and impact of immersive virtual reality therapy in patients hospitalized with decompensated cirrhosis. The study demonstrated that a single, guided virtual reality session was safe, well tolerated and effective in reducing anxiety and pain, while also enhancing patient engagement and overall well-being during hospitalization.
Publication
Digital Clinic for Alcohol-Associated Liver Disease (DALC)
DALC is a randomized, pragmatic clinical trial designed to evaluate the impact of a multidisciplinary, digitally delivered care model for people with alcohol-associated liver disease and alcohol use disorder. The primary aim is to determine whether integrating this digital approach into routine care improves patient outcomes, reduces healthcare use and enhances healthcare professional satisfaction compared with standard care.
Digital phenotype in diagnosis and prediction of outcomes
In collaboration with the Massachusetts Institute of Technology, this study leverages a passive, radio signal-based device to define a digital phenotype in patients with cirrhosis and evaluate its association with clinical outcomes.
AI is applied in various domains of care for patients with liver disease.