AI could help spot early Alzheimer's discoveries

According to the researchers, artificial intelligence (AI) could soon be used to detect Alzheimer's disease at the very beginning of its evolution. The results of the study were published in the last issue of the journal Radiology.

Researchers at the University of California at San Francisco conducted a small but significant study in which they used a self-learning computer program to examine participants' brain scans. The computer program was designed to detect the first signs of a neurodegenerative disease – Alzheimer's disease. These features were generally too minimal and subtle for human eyes to spot. The results revealed that the program could detect the signs of Alzheimer's disease in 40 patients six years (on average 75.8 months) before symptoms of the disease manifested and the disease was officially diagnosed.

The team wrote, "It is widely recognized that deep learning can help cope with the increasing complexity and volume of imaging data, as well as the diverse skills of imaging specialists. The application of machine learning technology to complex outcome patterns, such as those found in functional PET imaging of the brain, is only beginning to be explored. We hypothesized that the deep learning algorithm could detect features or patterns that are not evident during standard clinical examination of images and thereby improve the final diagnostic classification of individuals. "

The in-depth learning algorithm developed by the team was formed by showing 2109 F-FDG-FDG (or fluorine 18 (18F)) positron emission tomographies on 1002 patients. These PET scans can detect the metabolic activities of all parts of the brain by showing the absorption of a radioactive glucose compound administered by injection into the bloodstream. The program could detect changes in metabolic patterns of use in the brain that may signal the future development of Alzheimer's disease. The algorithm is driven by observing these changes in all the analyzes presented. Once they knew what they were looking for, they were able to detect all the first cases of the pilot study, the researchers explained. The pilot study used 40 PETs from 40 patients whose analysis had not yet been met by the algorithm. He predicted the first signs of the disease with 100% accuracy, the team wrote. Dr. Jae Ho Sohn, one of the co-authors of the study, said, "We are very pleased with the performance of the algorithm. He was able to predict every case that has progressed to Alzheimer's disease. "

The team explained that with this early detection, interventions and treatments could also start early. This could not only slow the progression of the disease, but also halt the progression of degeneration.

Dr. Carol Routledge, of the Alzheimer's & UK Research Association, has explained that these neurodegenerative disorders usually begin up to two decades before symptoms occur. This algorithm can help detect the disease at an early stage, so treatment can begin earlier, before many injuries are caused, she added. "This study highlights the potential of machine learning to aid in early detection of diseases such as Alzheimer's disease, but the results will need to be confirmed by much larger groups of people before we can properly assess the potential of this approach, "she said.