DeepSarc: A deep learning-based methodology for assessing sarcopenia and frailty in hospitalized older adults

Frailty is becoming more frequent for patients and the healthcare system as the Canadian population ages. Frailty is defined by a reduction in physiological reserves, which is linked to an increased risk of mortality. Sarcopenia, or age-related loss of muscle mass and function, is a key component of frailty that can be measured. Sarcopenia can also be treated in clinical settings to help older patients with acute and chronic illnesses to avoid the harmful implications of frailty. Unfortunately, diagnostic methods for sarcopenia are not always readily available or userfriendly for clinicians. Therefore, the proposed project establishes a novel methodology based on recently developed artificial intelligence techniques with the focus on older patients who have been admitted to the hospital with various forms of cardiovascular illness. This project will allow physicians to quickly assess their patients’ frailty, even while they are acutely unwell and bedbound, and guide decision-making so that they receive potentially beneficial interventions while avoiding those that are costly and futile.

Faculty Supervisor:

Hassan Rivaz

Student:

Partner:

Jewish General Hospital

Discipline:

Engineering

Sector:

Health and Related Sciences & Technology

University:

Concordia University

Program: