Alzheimer is a disease which affects people all over the globe and scientists have been trying for years to provide better treatment and to detect it early on. Early detection can help patients receive better treatment and slow down the evolution of the disease. In order to try and develop an early detection system, scientists are divided between groups which use blood and cerebrospinal fluid tests and groups developing gadgets to detect early sings. According to a team of researchers, AI can detect Alzheimer even 10 years early on.
AI study and methodology
The study was performed by a team of researchers from the University of Bari, Italy. The team developed an algorithm which is able to detect small structural changes in the brain caused by the disease even a decade before symptoms begin.
The methodology of the study involved training the AI by giving it 67 MRI scans (38 scans from Alzheimer’s patients and 29 scans from healthy patients). The scans were also divided into small regions and the AI analyzed the neuronal connectivity between. The second stage of the study involved testing the algorithm by having it process 148 scans of patients (48 scans were of patients who suffered mild cognitive impairment and developed afterward the disease).
Results and conclusions
The AI managed to diagnose the disease 86% of the time and detect mild cognitive impairment 84% of the time. This is very important because AI can become a tool used successfully for early detection.
The study also has limitations: the AI was given scans only from the USC LA’s Alzheimer’s Disease Neuroimaging database. This implies that further studies need to be conducted in which AI could be provided more samples so as to make the system even more accurate. The progress is promising.