Researchers at the University of Cambridge have developed an AI-based system that will diagnose dementia much faster than existing techniques using a single image.
Researchers at the University of Cambridge have managed to develop an artificial intelligence-based system that they believe can diagnose dementia from a brain scan, potentially years earlier.
Until now, the process of trying to determine whether a patient has dementia has been very complicated. It often lasts for long weeks, and involves subjecting the patient to a number of cognitive tests as well as several brain imaging examinations; Without being able to get a conclusive result each time.
On the other hand, the system proposed by the Cambridge researchers only needs to image what AI does best: spot patterns that are characteristic, but too abstract or complex for humans. It therefore makes it possible to achieve a conclusive result much faster; Important data in the management of these diseases.
In fact, many forms of dementia can experience a dramatic development shortly after the onset of symptoms; The sooner it is diagnosed, the more effective the treatment will be. This system also presents an added value, as it is also able to diagnose the course of the disease.
AI, the expected revolution in clinical science
Obviously, as always when it comes to health, the results have to be taken with a grain of salt. AI has been revolutionizing clinical science for some time now, but we are not yet ready to fully rely on it; At this point, it is imperative that humans remain in the loop. And for good reason: even though the AIs are much better than us at this game, they aren’t infallible. They completely depend on the dataset with which they are trained. It is therefore impossible to get absolute certainty through this method.
But that doesn’t mean AI has no place, far away from this area. This technology has recently entered the clinical trial phase; To judge its effectiveness, approximately 500 British patients This device should be tested in the coming year. In particular, it will allow to draw more clear conclusions on accuracy level of this technology. If found accurate enough, it can be a great diagnostic tool; Certainly enough to find a place of choice among specialists in neuropsychiatry and geriatrics. Incidentally, this is also a new example of the capabilities of AI applied to health, which could revolutionize our health system much sooner than we thought.