Smartphones and AI Breakthrough to Detect Alzheimer's

Overview

A machine learning model is being created by researchers with the goal of detecting Alzheimer's disease early. This model, which may be accessed via smartphones, can discriminate between people with Alzheimer's disease and healthy people with an accuracy of 70–75%. The technology might provide important early indicators by concentrating on speech patterns rather than content, thus starting therapy sooner and delaying illness development.

It won't take the role of medical experts, but it might improve telemedicine offerings and help people over linguistic and geographical hurdles. With an accuracy of 70–75%, the machine learning model can distinguish Alzheimer's patients from healthy people.

Instead of focusing on specific words, the program analyses linguistic and acoustic speech aspects to identify the ailment. The use of this model might lead to the creation of an easy-to-use screening software for smartphones that would show early signs of Alzheimer's disease. For the more than 747,000 Canadians who had Alzheimer's or another type of dementia, the model's accuracy in differentiating Alzheimer's patients from healthy controls was between 70 and 75 percent.

Early-stage detection of Alzheimer's dementia can be difficult since the symptoms frequently start off very subtly and might be mistaken for age-related memory problems. However, as the researchers point out, patients can start acting more quickly the earlier possible problems are identified. 

"Before, you'd need lab work, and medical imaging, to detect brain changes; this takes time, it's expensive, and nobody gets tested this early on," explains Eleni Stroulia, a professor at the Department of Computing Science who was involved in the model's development.

If mobile phones could provide an early signal of the patient-physician interaction, it would be helpful. It could enable us to begin the treatment sooner, and we might even begin by using mobile devices and small treatments at home to halt the development.

The Role of Healthcare Experts would not be Replaced by a Screening Tool

However, Zehra Shah, a master's student in the Department of Computing Science and the paper's first author, argues that in addition to helping with earlier detection, it would also create a practical way for patients who might experience linguistic or geographic barriers to accessing services in their area to identify potential concerns via telehealth. With this technique, Shah claims that it is possible to prioritize patients just based on speech.

The study team has previously studied the language used by Alzheimer's dementia patients, but for this investigation, they focused on general speech traits rather than words. "The initial process required hearing what the individual said, comprehending what they said, and grasping the significance. That computational issue is simpler to resolve, claims Stroulia". At this point, we advise listening to the voice. There are certain characteristics of human speech that go beyond language. Stroulia continues, "It's a lot more potent than the version of the problem we were solving before."

The researchers started with speech traits that medical professionals had noticed were typical of people with Alzheimer's disease. These individuals often spoke more slowly and interrupted their sentences more frequently. They frequently spoke with less clarity and tended to use shorter words. Researchers developed methods for converting these traits into speech cues that the model could test for. Despite concentrating on English and Greek speakers, Shah asserts that "this technology has the potential to be used across different languages." Although the concept itself is complicated, the final user experience for a product that uses it is really straightforward. 

According to Russ Greiner, a co-author of the article and professor in the Department of Computing Science, "A person talks into the tool, it does an analysis and makes a prediction: either yes, the person has Alzheimer's, or no they don't." A healthcare expert can then use that information to decide what the best course of action is for the patient.

The computational psychiatry research team at the University of Arizona is headed by Greiner and Stroulia, and its members have developed comparable AI models and tools to identify mental diseases like PTSD, schizophrenia, depression, and bipolar disorder. "Anything we can do to amplify the clinical processes, inform treatments, and manage diseases sooner with less cost is great," asserts Stroulia.

For more information do visit our journal’s homepage for once: https://www.pulsus.com/journal-clinical-psychiatry-neuroscience.html

For author submission: https://www.pulsus.com/submissions/clinical-psychiatry-neuroscience.html