Medical researchers have increasingly used artificial intelligence to analyze factors such as patients' sleep patterns, speech and typing to identify cases of early-stage dementia more quickly, according to a Nov. 2 report in The Wall Street Journal.
Below are four key studies on the subject.
- A National Institute on Aging-funded study published in 2020 in the Journal of the American Geriatrics Society analyzed data for more than 16,000 medical visits of 4,330 patients at Kaiser Permanente Washington health system. The research team developed a machine learning model that identified 31 factors linked with cognitive decline and was able to flag more than 1,000 visits that resulted in a patient being diagnosed with dementia. Nearly 500 of those diagnoses were for patients whose cognitive changes were previously undetected.
- In a 2020 study published in Current Alzheimer Research, researchers had nearly 8,900 people read short sentences aloud. Machine learning algorithms were able to accurately determine which participants were experiencing cognitive impairment and which were healthy by analyzing the acoustics of speech.
- A 2020 study in EClinicalMedicine examined 270 individuals' written speech patterns for signs of cognitive decline, accurately differentiated individuals experiencing cognitive decline from those who were healthy.
- A 2018 study published in NPJ DIgital Medicine analyzed the internet search behavior of 31,321,773 users. Researchers discovered that machine learning algorithms could examine users' cursor movements — including speed, tremors, direction changes and repetitive clicks — to help detect Parkinson's disease. Early analysis suggests the same approach could potentially detect Alzheimer’s as well.