Machine-learning model may help detect undiagnosed dementia

20 July 2018 (Last Updated July 20th, 2018 10:52)

A study led by the University of Plymouth in the UK has indicated the potential use of a machine-learning model to detect undiagnosed dementia in primary care.

A study led by the University of Plymouth in the UK has indicated the potential use of a machine-learning model to detect undiagnosed dementia in primary care.

During the study, the machine-learning model captured read-encoded data from 26,483 patients in participating National Health Service (NHS) general practice (GP) surgeries in Devon. The patients were more than 65 years of age.

Clinical and administrative data was used to train the machine-learning classification model to detect patients with signs of underlying dementia, considering factors such as weight and blood pressure.

“Machine learning is an application of artificial intelligence (AI) where systems automatically learn and improve from experience without being explicitly programmed.”

Findings demonstrated that the technology was able to correctly identify 84% of people with dementia and 87% of those that did not have the condition, which respectively indicated its sensitivity and specificity.

Based on these results, it is believed that the model could potentially decrease the number of people living with undiagnosed dementia.

University of Plymouth School of Computing Electronics and Mathematics professor Emmanuel Ifeachor said: “Machine learning is an application of artificial intelligence (AI) where systems automatically learn and improve from experience without being explicitly programmed.

“It’s already being used for many applications throughout healthcare such as medical imaging, but using it for patient data has not been done in quite this way before.

“The methodology is promising and, if successfully developed and deployed, may help to increase dementia diagnosis in primary care.”