Tuesday , December 1 2020

AI tells you when you can treat Alzheimer's disease: The program checks the brain for subtle signs long before the first symptoms



Artificial intelligence can help detect Alzheimer's dementia a few years before the first symptoms appear. In the first pilot study, the researchers trained learning algorithms to identify subtle signs of dementia in a special brain scan. This change in brain metabolism is often overlooked by doctors. However, the AI ​​system achieved a 100% hit rate, the researchers report.

Artificial intelligence helps detect early signs of Alzheimer's disease in the brain.

Artificial intelligence helps detect early signs of Alzheimer's disease in the brain.

One of the greatest advantages of AI is that learning computer systems are especially sensitive to the subtle similarities of data sets. Through extensive sample data training, you can study independently to identify melanoma in, for example, colon cancer precursors, breast cancer or medical imaging. In part, the hit rate of such a machine brain is already higher than the experienced physician.

Signed Brain Metabolism

But can artificial intelligence even interpret the signs that medical professionals generally do not miss? The classic case is early detection of Alzheimer's disease. In general, gradual destruction of brain matter is not diagnosed until the first symptom appears. Jae-Ho Sohn, a researcher at the University of California in San Francisco, said, "However, the loss of brain matter is too great to intervene.

Problem: There is a subtle change in brain metabolism that can be visualized using a special version of Proton Emission Tomography (PET). The uptake of radioactively labeled sugar molecules maps to other brain regions. However, typical changes in the early stages of Alzheimer's are difficult to detect. "We humans are good at finding specific biomarkers of disease," Hands explains. But these metabolic changes represent a subtle and diffuse process. "

Hitting rate 100%

Artificial intelligence works at this point. Son and his team wanted to know if the correct pattern recognition of learning algorithms could better detect Alzheimer's disease syndrome in brain scans than in humans. For their experiments, they trained PET images of Alzheimer's patients early in the AI ​​system. After that the AI ​​40 should not evaluate the recordings that have been seen before.

Amazing result: Artificial Intelligence has a 100% hit rate. "We were very pleased that this system could be verified every time we developed into Alzheimer's disease." Some of these PET images were taken several years before the first symptoms appeared and the final diagnosis began.

Early Diagnosis – Better Therapy

Researchers have shown that the AI ​​system can improve early detection of dementia. This was a preliminary test to be reviewed and supplemented by a more extensive multicentre study. Nevertheless, Sohn and his team are convinced that adaptive algorithms can support radiologists in the early detection of Alzheimer's disease in the future.

Early detection can help find new treatments. "Early detection of Alzheimer's disease will be an opportunity for researchers to find new and better ways to stop or slow the progression of the disease." The son says. (Radiology, 2018)


Source link