[ad_1]
PET images of Alzheimer's patients were initially used by thousands of people to train AI. (Photo: North American Radiation Society)
BerlinEarly detection is especially important in the fight against Alzheimer's disease. If you still have dementia that can not be cured, you can at least delay the process of getting medication.
"Diagnosis of Alzheimer's disease only when clear symptoms are present is too much for effective interventions," explains Son Ho-ho.
With a team at the University of California, San Francisco, the doctor has developed a new tool for early detection of Alzheimer's disease. It is an adaptive algorithm that can predict dementia until it is diagnosed by a doctor.
The researchers focused on the subtle metabolic changes in the brain caused by disease outbreaks. These changes can be visualized using imaging technology known as positron emission tomography (PET).
Seven factors favoring Alzheimer's disease
However, the signs of the early stages of the illness are so weak that they are hardly known to experienced physicians. "It is easier for humans to find specific biomarkers of disease," Sohn explains. "But metabolism is a much more subtle process."
The researchers used artificial intelligence training using data from the Alzheimer's Disease Neuroimaging Institute (ADNI). This data collection contains thousands of PET images of Alzheimer's patients in the early stages of the disease. 90% of these records were used to train algorithms and the remaining 10% were used to control success.
For the final test, AI finally had to analyze 40 images that were not submitted to her. The result describes the son as follows. "This algorithm was able to reliably detect all cases following the onset of Alzheimer's disease."
Physicians were impressed with the early detection of patients, in addition to a 100% hit rate. On average, this system has been symptomatic for more than six years before the actual diagnosis of the disease. "We were very happy with this result," said the son. However, the doctor knows that the test series is still relatively small and there are additional tests that need to confirm the results.
He found in algorithms the potential to be an important tool in the treatment of Alzheimer's disease. "If we can find the disease early, the researcher will have the opportunity to slow down the disease process or even find a better way to stop it."
Source link
