AI system diagnoses dementia before symptoms appear

Researchers in Britain, led by a Greek woman from the diaspora, have developed an innovative artificial intelligence system that can recognize the first signs of dementia in the brain before patients even have the first symptoms and before neurologists can make a diagnosis. The system, which does not need more than one imaging of the brain to draw conclusions, thanks to a special algorithm, has already begun to be tested clinically, although it is too early to say when it will be possible to use it in practice.

The algorithm has been “trained” with thousands of brain images of people with dementia, so it can then make comparisons when asked to “read” a new brain image. It is able to recognize early signs of dementia that can not “catch” the eye of even a specialist neurologist, thus paving the way for the detection of neurodegeneration of the brain in time, several years earlier than is possible today.

Researchers, led by Professor Zoe Kourtzis of the University of Cambridge and the Alan Turing Institute, who created the system, according to the BBC and the British Guardian and The Independent, are optimistic that this will be very helpful. As Dr. Kourtzi stated, “if we intervene in time, the treatments will be able to start earlier and slow down the progression of the disease, while at the same time a bigger brain damage will be avoided”.

Preclinical tests have shown that the system can diagnose dementia even when there is not the slightest obvious sign of damage to the brain. It can also predict whether dementia will remain stable for many years and develop sooner or later.

Clinical trials at Cambridge’s Adenbrook Hospital and other UK clinics, involving about 500 people and led by Cambridge-based neurologist Dr Tim Rittmann, will show how “the system” works in practice. As he said, the new “smart” system is “a fantastic development”.

Ms. Kourtzi studied Experimental Psychology at the University of Crete and received her PhD in Cognitive Neuroscience from Rutgers University in the USA. He did postdoctoral research at MIT and Harvard, specializing in Cognitive Neuroimaging, and then conducted research at the German Max Planck Institute for Biological Government and at the British University of Birmingham. Since 2013 she has been teaching Experimental Psychology at the University of Cambridge and is considered a specialist in Cognitive Computational Neuroscience, being a pioneer in the use of Artificial Intelligence in mind and brain health.

The development of algorithms for clinical use is a rapidly growing field internationally, “marrying” computer science with Neurobiology and Medicine. Various “smart” systems have already been developed for the early diagnosis – from imaging tests – of various diseases.