A new artificial intelligence system can predict the development of Alzheimer's disease with an accuracy of 100% and six years before the distinctive signs that physicians use to diagnose the disease.
There is no cure for Alzheimer's and the effectiveness of the best treatments for illness diminishes as it progresses.
This means early detection is the best single hope for the future of the 44 million people living with Alzheimer's all over the world.
Scientists at the University of California, San Francisco (UCSF) have taught AI how to detect subtle changes in the brain that would be imperceptible to the human eye.
To date, the system has provided a seemingly faulty evidence warning before the full onset of devastating Alzheimer's effects in a small sample of tests.
A new AI system can detect subtle markers of metabolic changes in the brain that predict the development of Alzheimer's – six years earlier and 100 percent accuracy. The system may recognize slight decreases in gray matter (A) in patients with Alzheimer's
Despite the fact that Alzheimer's are among the top 10 deadly diseases in the world, we have started to have a rudimentary understanding of this condition in recent years.
While we better understand the indicators and causes, Alzheimer's disease – such as football-related cerebrovascular disease, CTE – can not be definitively diagnosed until death even today.
Doctors simply use the removal process to exclude as many other causes of memory loss as possible.
This is done through memory assessments and assessments of any potential behavioral changes, as well as laboratory tests and brain imaging.
Most people have only four to eight years to live until they are diagnosed with Alzheimer's disease.
There are drugs available to help patients manage memory loss – such as Aricept, Namenda and Excelon – but they are not a fixed solution, they will not stop or reverse memory loss, they could simply slow progress and help make day with a day of living a little easier to administer.
Such medicines do not work or work just as well for everyone, and after a certain stage they do not work at all.
The more the Alzheimer's patient can initiate earlier, the better, but before doctors prescribe the medication, there is sufficient clinical evidence of the disease.
But AI may be the key to detecting Alzheimer's early enough to keep a memory function – for six years longer, the new study, published in Radiology, suggests.
The most obvious sign of the disease is the accumulation of beta-amyloid protein plaques in the brain.
But recently, scientists linked some metabolic changes to Alzheimer's disease.
It is not clear why, but the brains of those suffering from Alzheimer's and diabetics do not absorb and process glucose – our main source of energy – in the same way as the healthy brain.
This metabolic change happens earlier, but differences in glucose uptake in the brain are very subtle and diffuse, "said study co-author Dr. Jae Ho Sohn, radiologist at UCSF.
People are good at finding disease-specific biomarkers, but metabolic changes are a more global and subtle process.
So while the doctor could see the signs of a change in disparate images, it would take a long time not long for people to change these changes and their subtle patterns to sufficient patients over a sufficiently long period of time , to say with certainty what Alzheimer's predictions change.
But this is exactly the type of task that artificial intelligence intends.
Specifically, the new system's machine learning system has managed to learn models in nearly 2,000 brain scans taken from 1,000 patients.
Dr. Sohn, lead author of the study, Dr Benjamin Franc and their team tested the system on 40 scans that were completely new to the algorithm.
The AI system detected every patient who eventually developed Alzheimer's brain scans, taken on average six years before a person's diagnosis was made.
"We were very pleased with the performance of the algorithm," Dr. Sohn said.
There will be some time – and more tests – before the system is publicly available. But Dr. Sohn and his team have great hopes for the potential of AI.
If we diagnose Alzheimer's disease when all the symptoms have occurred, brain loss is so significant that it is too late to intervene, "he said.
"If we can detect it earlier, this is an opportunity for investigators to find better ways to slow down or even stop the disease process."