Tuesday , October 19 2021

AI recognizes Alzheimer's disease already prior to diagnosis


Shots of human brains using PET

Thousands of PET images of Alzheimer's patients at an early stage used the researchers to train their AI. (Photo: Radiological Society in North America)

BerlinIn the fight against Alzheimer's disease, early detection is particularly important. If the unbearable dementia is discovered early, it can at least slow down its course with medication.

"If we only diagnose Alzheimer's disease when there are clear symptoms, loss of brain volume is so high that it is usually too late for effective intervention," explains Jae Ho Sohn.

Together with his team from the University of California, San Francisco, the doctor has developed a new tool for early detection of Alzheimer's disease: an adaptive algorithm that reliably predicts dementia disease before a doctor's diagnosis.

The researchers focused their development on subtle metabolic changes in the brain caused by the onset of disease. Such changes can be visualized using a imaging technique known as positron emission tomography (PET).

Seven factors that benefit Alzheimer's

However, the traces of early stages of the disease are so weak that they are hardly recognizable even for experienced doctors. "It's easier for people to find specific biomarkers of disease," explains Sohn. "But metabolic changes are much more subtle processes."

The researchers educated their artificial intelligence using Alzheimer's Disease Disease (ADNI) data. Among other things, this data collection contains thousands of PET images of Alzheimer's patients in the very early stages of the disease. 90 percent of these recordings, the researchers used the training algorithm, the remaining 10 percent to control success.

For the final test, AI must finally analyze 40 images that had not been sent to her until then. The result describes the zone as follows: "The algorithm was able to reliably detect all cases that occurred later in Alzheimer's disease."

In addition to the 100 percent hit rate, physicians especially impressed the very early identification of the cases. On average, the system again recognized the symptoms more than six years prior to the disease diagnosis. "We were very excited about this result," says Son. However, the doctor knows that the test series was still relatively small and further testing has to confirm the result.

Nevertheless, in his algorithm he sees the potential of an important tool in Alzheimer's treatment: "If we can detect the disease earlier, it will allow researchers to find better ways to slow or even stop the disease process."

Source link