It is usually difficult and expensive to detect Alzheimer’s at an early stage, but with mobile technology it can be done

Researchers at the University of Alberta have developed a machine learning model for the early detection of Alzheimer’s dementia. This model can distinguish between Alzheimer’s patients and healthy individuals with an accuracy of 70-75% and can possibly be accessed through smartphones.

By focusing on speech patterns instead of content, the tool can offer invaluable early indicators that can potentially start earlier treatment and slow the progression of the disease.

While not a substitute for healthcare professionals, it can strengthen telehealth services and help overcome geographical or language barriers.

Can Detect Alzheimer’s with 75% Accuracy

The machine learning model can identify Alzheimer’s patients with an accuracy of 70-75% compared to healthy individuals.

The tool analyzes acoustic and linguistic speech functions instead of specific words to detect the disease.

The application of this model could be a simple, accessible screening tool on smartphones that provides early indicators of Alzheimer’s.

Usually Difficult and Expensive to Detect at An Early Stage

Alzheimer’s dementia can be difficult to detect at early stages because the symptoms are often rather subtle and can be confused with memory-related problems typical of advanced age. But as the researchers note, the earlier potential problems are discovered, the sooner patients can begin to act.

“In the past, you had to have laboratory tests and medical imaging to detect changes in the brain. It takes time, it’s expensive, and no one gets tested that early,” says Eleni Stroulia, a professor in the Department of Computer Science who was involved in creating the model.

Although the model itself is complex, the final user experience for a tool that includes it would be incredibly simple. A screening method would not replace healthcare professionals, but in addition to helping with earlier detection, it would create a practical way to identify potential concerns via telehealth services.