The AI solution for diagnosing early cognitive decline leading to Alzheimer’s disease
One of the biggest challenges facing Alzheimer's and dementia researchers today is finding a way to clearly identify patients suffering from the very earliest stages of cognitive decline. New research is suggesting that artificial intelligence could be the key to accurately predicting those patients most at risk of developing Alzheimer's.
With the majority of new potential Alzheimer's disease treatments failing in various stages of human trials, many researchers are turning to prevention, instead of cure, as perhaps the most effective way to combat the condition. However, the problem is that there currently is no effective way to diagnose a patient suffering from the earliest stages of cognitive decline associated with Alzheimer's.
Blood tests, PET scans, eye tests, genetics, and even sniff tests, are all being investigated as ways to identify the earliest stages of cognitive decline, but nothing has proved wholly effective at this point. A new study is now suggesting that an AI algorithm, trained to evaluate a variety of diagnostic data, could be effective at predicting whether a person is at an early stage in the disease, and if they are likely to significantly deteriorate over the following five years.
"At the moment, there are limited ways to treat Alzheimer's and the best evidence we have is for prevention," says Mallar Chakravarty, a computational neuroscientist working on the project. "Our AI methodology could have significant implications as a 'doctor's assistant' that would help stream people onto the right pathway for treatment. For example, one could even initiate lifestyle changes that may delay the beginning stages of Alzheimer's or even prevent it altogether."
The algorithm, trained on data from over 800 subjects, incorporates an assortment of biomarkers from MRI imaging to genotype and clinical information. The subjects studied varied from healthy senior citizens to those with complete, clinically apparent Alzheimer's disease. A small portion of subject data also offered up to six years of individual clinical information allowing the algorithm to generate an understanding of disease progression. This helped the system better predict the trajectory of cognitive decline and it relates to these biomarkers.
At this early stage the researchers are confident the algorithm is robust, accurate, and would be useful as a tool for clinicians in better designing potential preventative treatments. But, as more data points are added to the algorithm, and it is trained on larger volumes of patients, the team is confident it will deliver even better predictions.
"We are currently working on testing the accuracy of predictions using new data," says Chakravarty. "It will help us to refine predictions and determine if we can predict even farther into the future."
The new study was published in the journal PLOS Computational Biology.
Source: McGill University