BUSINESS INSIGHTS : The latest news, analysis, and trends about protection and care

Feb 14, 2019,  by Allianz Partners Business Insights

Alzheimer’s: Artificial intelligence can diagnose the disease six years before doctors

One of the many potential applications of artificial intelligence in the field of medicine, AI could prove particularly useful in the early diagnosis of Alzheimer’s disease. Thanks to AI, the disease can be detected an average six years earlier than the traditional clinical approach, say American researchers.

Photo credits: Geralt/Pixabay

Canadian medical journal Radiology has just published a highly promising study by the University of San Francisco about the use of artificial intelligence (AI) in detecting Alzheimer’s, reports PourquoiDocteur. Scientists have managed to teach an AI how to identify the first symptoms of the disease, thanks to “deep learning”.

The tool, after having analysed a large number of brain scans, was able to spot the early signs of the disease an average of six years before a traditional clinical analysis could be carried out. This would save precious time in patient treatment.

More than 90% successful

Researchers from the University of San Francisco first collected 2,000 images from brain PET scans (3D medical imaging) of 1,002 patients. These images show up the quantities of certain key molecules for diagnosis, including glucose.

Glucose levels vary according to how active the brain’s cells are. The less active the brain cell, the less glucose it consumes. These variations are symptomatic, but are too small to be diagnosed by man. 

The machine created by the researchers has been successfully tested. It analysed a first series of 188 scans, then a second set of 40, achieving an impressive 92% and 98% success rate. 

Early diagnosis for better treatment

Though there is still no cure for Alzheimer’s, there are many medical treatments that can slow down its development. However, they are only effective if they are administered from when the first symptoms appear. Early treatment is currently quite difficult to put in place, but it could be made easier if diagnoses were made using AI.   

It remains to be seen when the system developed by the University of San Francisco, which is still in the test phase, might be used in hospitals.

Allianz Partners

Linked Topics
About Allianz Partners Business Insights