[ad_1]
It is very important to be able to diagnose diseases as serious as Alzheimer's disease, because treatments and interventions tend to be more successful in the early stages of the disease. However, early diagnosis seems to challenge the physicians themselves. The researchers have now established a link between disease progression and metabolic changes, which shows that some changes in glucose uptake in certain areas of the brain could be related to the disease, although these changes are difficult to recognize.
"The differences in glucose profiles in the brain are very subtle and diffuse" said co-author of a study, Jae Sohn, of the Department of Radiology and Biomedical Imaging at the University of California at San Francisco. "People are good at finding specific biomarkers, but metabolic changes are much more global and subtle processes," he said.
The main author of the study, Dr. Benjamin Franc, from the same university, helped by student Yiming Ding via the idea of the big data research group in radiology (Big Data in Radiology – BDRAD), consisting of doctors and engineers, They focused on the science of radiological data. Dr. Frank was interested in deep learning, a type of artificial intelligence in which machines learn by the example that they feed, as do human beings, so that this technique can be applied to detect changes in the metabolism of animals. people end up predicting Alzheimer's disease.
The researchers formed the deep learning algorithm through a technology called FDG-PET: 18-F-fluorodeoxyglucose positron emission tomography, in which the radioactive glucose compound is injected into the bloodstream. Positronic analysis can measure the uptake of FDG in brain cells, which are an indicator of metabolic activity.
The work was based on data from ADNI – Alzheimer's Disease Neuroimaging Initiative, site devoted to clinical trials aimed at improving the prevention and treatment of Alzheimer's disease. The ADNI dataset included more than 2100 FDG-PET images from 1002 patients. The researchers then trained the neural network in depth learning with 90% of the dataset and tested it with the remaining 10%. Thanks to an in-depth learning, the algorithm was able to self-learn the metabolic patterns corresponding to the terrible disease.
Finally, the researchers tested the algorithm in an independent set of 40 images of 40 patients that had never been studied before. The algorithm was 100% sensitive to detect the disease more than six years before its final diagnosis.
"We are very pleased with the performance of the algorithm," Sohn said. "He has been able to predict all the cases that have evolved into Alzheimer's disease."
We have to say that we have to be careful because the dataset we worked with is not important enough. However, for Sohn, this algorithm could prove useful as a tool complementing the radiological work, especially in conjunction with other imaging tests, thus enabling early therapeutic intervention.
"If we can diagnose Alzheimer's just when all the symptoms have occurred, it's hard to intervene," says Sohn. "On the other hand, if we can detect the disease early, there is an opportunity to treat the disease and even stop it."
[ad_2]
Source link
