Artificial Intelligence-Enabled Stool Analysis for Lactulose Titration Assistance in Hepatic Encephalopathy Through a...

Area of Science:

• Hepatology
• Artificial Intelligence
• Digital Health

Background:

• Hepatic encephalopathy (HE) management often involves self-titration of lactulose, which can be challenging for patients.
• Objective tracking of bowel movements and stool consistency is crucial for effective lactulose titration in HE.

Purpose of the Study:

• To assess the feasibility and effectiveness of an AI-enabled smartphone application for guiding lactulose use in patients with HE.
• To evaluate the application's ability to objectively track bowel movements and Bristol Stool Scale (BSS) for optimizing HE treatment.

Main Methods:

• A feasibility trial involving subjects with HE on lactulose.
• Subjects captured daily bowel movement pictures via a smartphone app.
• The app provided daily feedback on lactulose titration based on bowel movement frequency and BSS.

Main Results:

• The AI tool demonstrated high user satisfaction (>80%).
• Less compliant subjects significantly improved BSS goal achievement during the intervention phase (56% to 79%, P=0.041).
• The severe/recurrent HE group showed improved BSS goal achievement (77% to 82%), while the stable group showed no significant difference.

Conclusions:

• The Dieta application shows promise as a tool for objective Bowel Movement/Bristol Stool Scale tracking in hepatic encephalopathy.
• This AI-enabled tool may assist in optimizing lactulose titration for improved HE management.