AI can detect Alzheimer's ten years before the onset of the disease

New developments could make it possible to predict Alzheimer's early
Alzheimer's disease affects more and more elderly people today. Doctors have been looking for ways to determine the risk of developing the disease early on. Researchers from Italy have now developed an artificial intelligence that Alzheimer's can recognize ten years before the first symptoms develop.

Scientists at the University of Bari in Italy developed software that can detect small structural changes in the brain that years later can lead to mild cognitive disorders and the development of Alzheimer's. The doctors have now published the results of their study.

Alzheimer's disease prediction was 86 percent accurate
The newly developed artificial intelligence for the early detection of Alzheimer's, analyzed the images from a magnetic resonance tomograph and was able to diagnose Alzheimer's with a probability of 86 percent, ten years before the first symptoms appear, the researchers report.

Artificial intelligence detects changes in the brain
Small structural changes in the brain can indicate cognitive disorders and Alzheimer's. The algorithm of the new software can detect such changes and thus determine the likelihood of developing Alzheimer's within the next ten years, the authors explain. Artificial intelligence looked for differences in the connectivity of the brains of later Alzheimer's patients and the control group.

Early diagnosis makes it easier for patients to deal with the disease
If Alzheimer's is diagnosed early, it is certainly not a cause for joy for those affected. However, this early diagnosis enables patients to better prepare for the future disease. In this way, those affected can regulate family and financial matters at an early stage. In addition, it will make it possible for doctors to counteract the effects that will arise in the future and thus significantly slow the development of Alzheimer's disease.

Software detects Alzheimer's and cognitive disorders with high accuracy
A total of 67 scans from magnetic resonance tomographs were analyzed during the examination. 38 of the scans were from Alzheimer's patients, the other 29 scans were from a control group. When examining the data, the artificial intelligence detected Alzheimer's disease with an accuracy of 86 percent, the researchers explain. The software recognized slight cognitive disorders with an accuracy of 84 percent, according to the result of the evaluation of a further data set of 148 MRI scans.

Analysis of larger data sets could further improve accuracy
The scans used in the investigation came from the University of Southern California at Los Angeles. Larger amounts of data could further improve the training of artificial intelligence, the researchers explain. This would lead to increased accuracy in diagnosis. The scientists emphasize that artificial intelligence could enable even more reliable early detection of Alzheimer's.

Future treatments for Alzheimer's disease
In Alzheimer's disease, indissoluble plaques of the beta-amyloid protein fragment develop in the brain. Researchers are currently trying to develop a vaccine and medication to treat Alzheimer's early, which can either slow the disease down or, in the best case, even stop it completely, years before Alzheimer's symptoms break out. Vaccination could, for example, lead to antibodies against beta-amyloid developing preventing their deposition. The drug is said to block the enzymes that are involved in the development of Alzheimer's. (as)

Author and source information

Video: Artificial Intelligence Tool Developed for Alzheimers Research (June 2021).