Alzheimer's disease refers to an incurable neurodegenerative disease that develops asymptomatically in the body for decades. Currently, the disease is most often diagnosed already at an advanced stage, when a person has a cognitive decline. Early diagnosis can slow down this process, but there is no simple and universal screening tool yet.
One of the main diagnostic methods is MRI, which will determine the signs of the disease from images of the brain. Meanwhile, the results do not always provide accurate data, so (in addition to being examined by a doctor and passing cognitive tests), it is desirable to compare them with indicators from other tools, such as a study of cerebrospinal fluid.
Now, scientists from Kaunas University of Technology have unveiled an algorithm that can diagnose Alzheimer's disease from MRI scans with 98% accuracy.
These figures are not final - scientists are constantly improving the capabilities of AI through continuous analysis of data from new patients. It is important to note that the features of the algorithm do not depend on the equipment used in the hospital, so the technology should become a universal diagnostic tool in the future.
The ultimate goal is to develop a model that can label diseased areas of the brain for physicians to provide a visual picture of the progression of neurodegeneration. While the authors do not report when the technology will be available in clinical practice.
Recently, scientists from France published the main signs that, 15 years before symptoms, can signal the risks of developing an incurable disease.
No comments:
Post a Comment