July 2, 2020

They design a technique that generates a brain map to diagnose Alzheimer's

They design a technique that generates a brain map to diagnose Alzheimer's

A team of researchers from the University of Granada has designed an image processing technique that generates a spherical map of the brain, a tool that converts the information of resonances into a drop-down and that allows diagnosing Alzheimer's with 90% accuracy.

One of the researchers of this project of the Department of Signal Theory, Telematics and Telecommunications of the University of Granada, Francisco Jesús Martínez-Murcia, explained to Efe that Cerebral Spherical Mapping is a technique that allows analyzing and visualize medical images in an unprecedented way until now.

This advance has consisted in processing magnetic resonances of the brain and construct statistical maps that analyze the texture of each zone thanks to coordinate vectors that allow to select each voxel -like a pixel but volumetric- in all the dimensions of the brain.

In this way, according to Martínez-Murcia, a kind of map is generated in two dimensions that can be deployed to analyze the textures and diagnose diseases that cause cognitive deterioration such as Alzheimer's.

"One of the novelties is that it modifies the visualization method for physicians, because it presents a map that is easier to interpret than current ones in three dimensions and that shows textures and the areas affected by this type of disease," he said. investigator.

To verify the effectiveness of this brain map, which allows to check how the tissues vary in each direction or their densities, researchers have analyzed the resonance database of around a thousand people affected by Alzheimer's.

These tests have shown that the new mapping technique allows diagnosing Alzheimer's with more than 90% accuracy, since it demonstrates the decrease in tissue density caused by the process of neurodegeneration of the disease.

The two-dimensional visualization allows easy identification of the areas that contribute the most to differentiating between Alzheimer's patients and healthy subjects, located mainly in areas such as the hippocampus or the amygdala.

"But perhaps even more important is that we have achieved a diagnosis of 77% in a much more serious problem, to predict whether patients who already suffer mild cognitive impairment are going to progress to more advanced stages of Alzheimer's disease or will remain stable for several years. years, "the researcher stressed.

This technique also has an open source and is available for download for free in the "github" of the team of researchers.


Source link