Geisinger, Eisai to Use Artificial Intelligence to Detect Dementia

Geisinger and Eisai plan to study the potential of artificial intelligence (AI) by testing the effectiveness of an algorithm to identify individuals with cognitive impairment who may develop dementia.

According to Alzheimer’s Disease International’s World Alzheimer Report 2021, there are currently 55 million people worldwide living with dementia, and this is likely to increase to 78 million by 2030.

Using Geisinger’s deidentified patient dataset, the organizations will test the abilities of an AI algorithm called Passive Digital Marker (PDM) to define cognitive impairment, which can lead to the identification of dementia-related diseases such as Alzheimer’s. Researchers from Purdue University and Indiana University created PDM, which is trained using structured and unstructured data from three EMR datasets: diagnosis, prescriptions, and medical notes.

“As we continue to develop new treatments to prevent and slow the progression of Alzheimer’s and related dementias, early detection is becoming even more important,” said Glen Finney, MD, director of Geisinger’s Memory and Cognition Program and a board member of the Greater PA Chapter of the Alzheimer’s Association, in the press release. “Early and accurate diagnosis and treatment of these conditions can drastically improve outcomes and quality of life for both patients and caregivers.”

Generally, 40 to 60 percent of adults with probable dementia do not receive a diagnosis. Researchers believe that the implementation of AI for detecting the disease will be quick and efficient, allowing for early diagnosis and increased time for treatment.

“AI technology has the potential to transform medicine,” said Yasser El-Manzalawy, PhD, principal investigator and assistant professor of translational data science and informatics at Geisinger, in the press release.

“AI-based tools can efficiently scan massive amounts of healthcare data and identify hidden patterns. These patterns can be used to detect diseases, like cancer and dementia, at an early stage. Our data science research team is uniquely positioned to leverage this innovative technology to develop and validate tools to identify patients with unrecognized dementia or patients at high risk of developing dementia in the future,” said El-Manzalawy.

Previous research has shown using AI to identify patients with Alzheimer’s Disease can be a successful strategy. In August 2020, Stevens Institute of Technology researchers created an AI tool that diagnosed Alzheimer's with over 95 percent accuracy. The tool evaluated neural networks and identifed signals commonly linked to Alzheimer’s Disease.

Another area in which AI is being used to improve diagnoses is heart care.

In a study published in August 2020, researchers used an AI-enhanced electrocardiogram (ECG) to detect heart failure. Researchers developed this system using extensive patient data and training computers to define the differences between ECG patterns that do and do not indicate left ventricular systolic dysfunction.