Can AI provide early detection of delirium in patients with dementia? Bouvé researchers to study

10/02/19 – BOSTON, MA. – Artificial Intelligence stock on Oct. 2, 2019. Photo by Matthew Modoono/Northeastern University

Can artificial intelligence, such as home health monitoring devices and wearables, detect delirium in older persons with dementia before it clinically manifests? Two Bouvé researchers have been granted funding to learn more.

Jane Saczynski, professor in the Department of Pharmacy and Health System Sciences, and Carla Bouwmeester, clinical professor in the Department, have been granted funding from the NIH AI dementia center for their project.

According to Saczynski and Bouwmeester, delirium, an acute confusional state, is common in older adults with dementia and reflects underlying acute illness, but how can it be detected early?

They say home monitoring devices and wearables offer great promise for early identification of health changes that often lead to delirium – such as sleep disturbances, infection, and dehydration – but the tools have not been used to detect delirium in persons with dementia.

The team will work with clinical partner Harbor Health Elder Service Plan, a Program for All Inclusive Care of the Elderly (PACE), where nearly half of patients have a dementia diagnosis. Fifty patient/caregiver dyads will be enrolled in the study and their homes will be outfitted with remote health monitoring devices for eight months. These devices will monitor the patients’ sleep quality/disturbances, hydration, temperature, activity, blood pressure, and heart rate.

The research team believes this will have an immediate clinical impact on patient care in the Harbor Health PACE program and help prevent emergency room visits and hospitalizations. The data could then allow for larger clinical trials in the 30 states that have PACE programs.