Hepatic encephalopathy (HE) is a well-known complication of end-stage liver disease (ESLD), marked by a wide spectrum of neuropsychological abnormalities from accumulation of toxins in the brain. HE is the second most preventable cause of hospital admissions for patients with ESLD, with estimates of 40% for 30-day readmissions and is responsible for approximately $7.2 billion in hospital charges in 2009, a number that continues to rise every year. HE is associated with high morbidity and mortality (15% inpatient mortality) if allowed to progress to its later stages. Therefore, early detection and intervention of HE are paramount in halting its progression and preventing complications. However, early stage HE is very difficult to detect in routine clinical settings given limited known diagnostic modalities. Existing tools are cumbersome to use, often limited to research settings, and are only validated for detection but not monitoring of HE progression, the latter of which is often more clinically useful. Therefore, we propose a novel diagnostic and monitoring tool that utilizes a combination of speech, movement, and cognition testing modalities trained using machine-learning that are sensitive and specific enough to both diagnose and grade early stage HE. The tool will be designed to be quick and user-friendly to administer both in the clinic and at home, with the ultimate goal to improve outpatient HE management, reduce unnecessary hospital admissions and associated costs, as well as improve the quality of life of patients with tailored, individualized therapies.

Team Lead 
Xing Li

Massachusetts General Hospital, Boston

Massachusetts General Hospital Medical Innovation Program

Lead Mentors
Jason Tucker-Schwartz
Claire Zhao