Doctors are working hard to detect Alzheimer’s disease sooner and sooner, but it looks like artificial intelligence has them beat. Researchers have developed an AI that is able to detect the neurodegenerative disease – which leads to loss of memory and cognitive functions – almost 10 years earlier than doctors can by looking at symptoms alone.
Using MRI scans, the AI can detect early signs with 84 percent accuracy by identifying changes in how regions of the brain are connected.
‘Nowadays, cerebrospinal fluid analyses and brain imaging using radioactive tracers can tell us to what extent the brain is covered with plaques and tangles, and are able to predict relatively accurately who is at high risk of developing Alzheimer’s 10 years later,’ Mariana La Rocca, a researcher on the study, told New Scientist of the current detection methods.
‘However, these methods are very invasive, expensive and only available at highly specialised centres.’
One of the main benefits of the new AI is that it’s noninvasive, simpler and cheaper
La Rocca along with Nicola Amoroso and other colleagues at the University of Bari in Italy developed the machine learning algorithm and found it was able to effectively understand the changes in the brain related to Alzheimer’s.
They used 67 MRI scans -38 of which were from people who had Alzheimer’s and 29 from healthy controls – to train the algorithm to distinguish between healthy brains and those at-risk for the disease.
The researchers then divided the brain scans into small regions in order to analyze the neuronal connectivity between them, making no assumptions about the diagnoses.
They tested the AI on a set of scans from 148 subjects, of which 52 were healthy, 48 had Alzheimer’s disease and 48 had mild cognitive impairment (MCI) but were known to have developed Alzheimer’s disease 2.5 to nine years later.
It was able to tell the difference between a healthy brain and one with Alzheimer’s with 86 percent accuracy.
The AI also proved successful in recognizing the symptoms: it was able to successfully distinguish between healthy brains and those with MCI with 84 percent accuracy.
Doctors aren’t able to identify the disease until a decade past the MCI stage.
They found the AI was most effective when the brain regions being compared were about 2250 to 3200 cubic millimetres in size.
‘The information content provided by multiplex characterization (AI) was able to efficiently detect disease patterns,’ the paper concludes.
‘Also, the method is very suitable to application to longitudinal studies, ideally in association with functional imaging, to improve our understanding of the different patterns of neurodegeneration in different diseases.’
Other researchers have praised the findings, with Patrick Hof at the Icahn School of Medicine at Mount Sinai in New York saying it would be ‘incredibly valuable.’
Next, the team wants to work on apply the AI to other neurodegenerative disease – such as Parkinson’s diseases – as well.
‘It’s a method that’s very versatile,’ she said.
Daily Mail
UM– USEKE.RW
