[ad_1]
Telling Signs: Artificial intelligence could help detect Alzheimer's dementia years before the onset of the first symptoms. In a first pilot study, researchers formed a learning algorithm to identify subtle signs of dementia during special brain tests. These changes in brain metabolism are often difficult to recognize for physicians. However, the researchers report indicates that the AI system has achieved a 100% success rate.
Artificial intelligence can help detect the first signs of Alzheimer's disease in the brain.
© lagereek / thinkstock 
This is one of the great strengths of artificial intelligence: computer systems learning is particularly effective at detecting subtle similarities in data sets. Through training on many data samples, they can learn independently, for example, to identify precursors of colon cancer, bad cancer or melanoma on medical images. In part, the success rates of these brain machines are already higher than those of experienced physicians.
Connect to brain metabolism
But can AI also interpret omens that even health professionals usually miss? A clbadic case is the early detection of Alzheimer's dementia. As a general rule, the progressive deterioration of the brain material is only diagnosed when the first symptoms occur. "But the loss of brain substance is so severe that it's usually too late to intervene," said Jae-Ho Sohn of the University of California at San Francisco.
The problem: Although brain metabolism undergoes subtle changes, these can be visualized with the help of a special version of positron emission tomography (PET). The uptake of radioactively labeled sugar molecules is mapped in different regions of the brain. However, typical changes in early stages of Alzheimer's are difficult to detect. "We humans are able to find specific biomarkers for the disease," says Sohn. "But these metabolic changes represent a more subtle and diffuse process."
100 percent success rate
At this point, artificial intelligence comes into play: His and his team wanted to know if the correct pattern recognition of a learning algorithm could possibly better detect the sign of the disease. Alzheimer's in the brain scanner that a human being. For their experience, they formed the AI system on about 2,000 PET scans of patients with early Alzheimer's disease. Then, the AI 40 should not evaluate the previously seen records.
The amazing result: artificial intelligence showed a success rate of 100%. "We were very happy: the system was able to identify each case that eventually turned into Alzheimer's disease," said Sohn. Some of these PET images were taken several years before the onset of the first symptoms and a definitive diagnosis.
Early diagnosis – best therapy
According to the researchers, this proves that AI systems could improve the early detection of dementia. Although it only involves a pilot study, it should be examined and supplemented by more in-depth and multi-center studies. Nevertheless, Sohn and his team are convinced that adaptive algorithms can help radiologists and neurologists to quickly detect Alzheimer's disease.
Early identification can also be helpful in finding new treatments: "Recognizing Alzheimer's disease earlier will be an opportunity for researchers to find new, more effective ways to stop or slow the progression of the disease." said the son. (Radiology, 2018)
Source link
