3Comments
An artificial intelligence (AI) system has been developed that can identify the early markers of Alzheimer’s disease (AD) with over 99 percent accuracy. By assessing brain scans of older adults, the algorithm is able to pick out subtle changes that often occur before diagnosis, thereby enabling doctors to provide early treatment to high-risk individuals.
In the journal Diagnostics, the study authors explain how their AI successfully recognizes signs of mild cognitive impairment (MCI), considered an intermediate stage between the expected cognitive decline associated with normal aging and AD. While MCI typically produces no noticeable symptoms, it is linked to changes in certain brain regions that can be detected on functional magnetic resonance imaging (fMRI) scans.
However, manually searching for these changes can be tricky, and doctors don’t always spot them when looking at scans. By repurposing an existing neural network called ResNet18, the researchers created an AI model capable of identifying MCI with greater reliability.
“Modern signal processing allows delegating the image processing to the machine, which can complete it faster and accurately enough,” explained study author Rytis Maskeliūnas in a statement. “Of course, we don’t dare to suggest that a medical professional should ever rely on any algorithm one hundred percent.”
To create their AI, the researchers trained the neural network on 51,443 brain scans from 138 people. These images fell into six distinct categories, ranging from healthy brains through to various degrees of MCI and full-blown AD. A further 27,310 images were then used to validate the algorithm, which was able to identify early MCI with 99.99 percent accuracy and late MCI with 99.95 percent accuracy.
“The proposed model performed better than other known models in terms of accuracy, sensitivity, and specificity,” write the authors, adding that their system is “more trustworthy and accurate” than existing diagnostic tools for future Alzheimer’s risk.
Importantly, the researchers stress that MCI doesn’t always lead to AD, and that individuals who show signs of these brain changes may not necessarily go on to develop the condition. However, identifying MCI enhances the ability of healthcare professionals to assess a patient’s risk for Alzheimer’s, potentially allowing for earlier screening and intervention.
Describing how the algorithm could be used in practice, Maskeliūnas explained that “after the computer algorithm selects potentially affected cases, the specialist can look into them more closely, and at the end, everybody benefits as the diagnosis and the treatment reaches the patient much faster.”
