Although Alzheimer's disease affects tens of millions of people around the world, it remains difficult to discover at an early stage. But researchers dealing with artificial intelligence capabilities in medicine have found that technology can help early diagnosis of treating ailments. The California team recently published a report on its research in the Radiologists and showed that a once-trained neural network was able to accurately diagnose Alzheimer's disease in a limited number of patients based on the visualization of the image of the brain done several years before the patients were diagnosed by the doctor.
The team uses brain imaging (FDG-PET imaging) to train and test their neural network. In FDG, the patient's bloodstream images are injected with a radioactive type of glucose, and then his body in the body, including the brain, pushes it towards the surface. Scientists and doctors can then use a PET scan to sensitize the metabolic activity of this tissue, depending on how much FDG is taken.
The FDG-PET method is used to diagnose Alzheimer's disease, where patients with disease usually exhibit lower levels of metabolic activity in certain parts of the brain. Experts, however, need to analyze these images to find evidence of disease, and this becomes very difficult because moderate cognitive impairment and Alzheimer's disease can lead to similar scans.
For this reason, the team uses 2,109 FDG-PET images of 1002 patients, trains their neural network to 90% and tests them for the remaining 10%. He also tests with a unique set of 40 patients scanned between 2006 and 2016, and then compares the findings of artificial intelligence with those from a group of specialists who analyze the same data.
With a separate set of data for testing, Artificial Intelligence is able to diagnose Alzheimer's patients with 100% accuracy and with 82% accuracy of those who do not suffer from treating a disease. It can also make forecasts on average more than six years in advance. For comparison, a group of doctors who looked at the same scanned images identified patients with Alzheimer's disease in 57% of cases and those with no disease – 91%. However, the differences in the performance of machines and people are not so noticeable when it comes to the diagnosis of mild cognitive impairment that is not typical of Alzheimer's disease.
Researchers note that their research has several limitations, including a small amount of test data and limited training data.