Another Step Forward in Alzheimer’s: AI Can Now Predict Its Possible Onset With 99% Accuracy

According to a recent article in Science Daily, the WHO lists Alzheimer’s disease as the cause of dementia in approximately seventy percent of cases worldwide or a total of twenty-four million afflicted people. That number is expected to double every twenty years.

Medical Professionals are Raising Awareness

The earlier we can manage to identify the onset of Alzheimer’s the easier it will be to benefit from emerging new treatments. Although the onset of Alzheimer’s usually does not indicate overt symptoms, neuroimaging can occasionally bring about early detection such as mild cognitive impairment (MCI). This is the stage between normal aging and dementia.

The Alzheimer’s Disease Neuroimaging Initiative

On a positive note, Modupe Odusami Ph.D. a researcher from the AI Sector of Lithuania’s Kaunas University created an AI deep learning algorithm obtained from MRI images of 138 subjects. These images were divided into six categories that ranged from healthy images to MCI and on to Alzheimer’s disease. The images were used for validation and training.

The algorithm is capable of predicting the onset of Alzheimer’s disease. The test performance was superior to previous test methods in terms of sensitivity, accuracy, and specificity which means the extent to which the tests applied to Alzheimer’s.

In addition, and according to previous research, fMRI (functional magnetic resonance imaging) has been used to recognize regions of the brain in connection with the early stages of Alzheimer’s. The procedure requires not only specific knowledge but is also extremely time-consuming. That is where AI and deep learning can take over and speed the process significantly.

A word of caution about discovering MCI features. This does not always mean that the symptoms are associated with Alzheimer’s. It may just be an indicator that an examination by another medical professional is advised. The researchers were clear that they would not suggest that anyone rely solely on an algorithm (problem-solving formula). Rather, they advise people to think of the mechanism as a robot that is sorting data.

Dr. Maskeliūnas supervised the project. He was pleased to announce that in the aforementioned scenario, everyone benefits. In other words, the diagnosis and the treatment are expedited. Dr. Maskeliūnas mentioned that the team had previously attempted to diagnose Alzheimer’s in its early stages, but this is the first time they could rely on an accurate algorithm.

Looking Forward

According to Dr. Maskeliūnas, software could be designed using the current algorithm that would analyze data collected from people over the age of 65 who have had an unusual medical history such as a brain injury or high blood pressure. Medical personnel would then be notified.

Dr. Maskeliūnas, explains that they work on the principle of European open science. Therefore, as a contribution to medical advancement, their findings can be used, expanded, and developed further.

The head of the research sector who specializes in the application of AI on signal processing adds that the new model may be integrated into an even more complex system that can monitor eye movement, voice analysis, or facial reading. He explains that this technology may be applied to a self-check so that the patient would seek advice from a professional if needed.

Both doctors agree that for the moment technology cannot replace the medical professional, but it will provide cheaper and more accessible medicine.

Rose Duesterwald

Rose Duesterwald

Rose became acquainted with Patient Worthy after her husband was diagnosed with Acute Myeloid Leukemia four years ago. He was treated with a methylating agent While he was being treated with a hypomethylating agent, Rose researched investigational drugs being developed to treat relapsed/refractory AML.

Share this post

Share on facebook
Share on google
Share on twitter
Share on linkedin
Share on pinterest
Share on print
Share on email