Alzheimer’s Disease is a terrible condition that impacts millions of people. While there is no cure for it, there are treatments available that can slow it down and help patients to improve their quality of life. As with almost any disease, detecting it early helps to provide patients and doctors with additional options when it comes to treatment and management.

With that in mind, there was an exciting study conducted by scientists at Boston University and published in Alzheimer’s & Dementia. The researchers came up with a new artificial intelligence (AI) algorithm that analyzed speech patterns of people who have mild cognitive impairment (MCI).

From there, the AI could predict which of those patients would progress to Alzheimer’s within six years and which would not. The AI had an accuracy rate of 78.5%, which is very impressive.

The study took recorded and transcribed audio recordings from 166 individuals with MCI and fed it into the AI to train the system. The recordings were collected from people who were 63-97 years old.

From there, the team fed in speech samples from patients that were recorded in the past. Since the doctors already knew which patients developed Alzheimer’s and which did not, they could evaluate the success rates of the AI immediately. Computer scientist, Joannis Paschalidis of Boston University said the following:

“You can think of the score as the likelihood, the probability, that someone will remain stable or transition to dementia. We wanted to predict what would happen in the next six years – and we found we can reasonably make that prediction with relatively good confidence and accuracy. It shows the power of AI.”

One of the many things that makes this study promising is that the audio that was fed into the system was not of especially good quality. Since the recordings were taken in the past and were made without the specific intent of being used in a medical study, they just offer a baseline of the possibilities of the technology.

In the future, higher quality recordings can be taken and fed in to hopefully get even more accurate results. Researchers may even be able to come up with an optimized set of things to record in order to make further improvements.

If the results continue to be as positive as those found in this study, it will provide people with an easy and non-invasive way to evaluate their risk of Alzheimer’s in the future. Some researchers even envision a day when people will be able to make recordings of themselves, possibly on an app on their phone, that can be used to track their speech patterns over time. Their doctors (or even just an AI system) can look at the results and provide feedback.

Early detection will also help future researchers to come up with better treatment options that could even lead to a cure.

If you thought that was interesting, you might like to read about the mysterious “pyramids” discovered in Antarctica. What are they?