Algorithm That Detects Brain Defects Could Help Treat Epilepsy
A multinational research team has developed an artificial intelligence algorithm capable of identifying minute brain anomalies that can lead to epileptic seizures.
The Multicentre Epilepsy Lesion Detection project (MELD) used more than a thousand patient MRI scans from 22 epilepsy centers around the world to develop the algorithm, which provides reports of where abnormalities are in cases of drug-resistant focal cortical dysplasia (FCD) – a leading cause of epilepsy.
FCDs are areas of the brain that have developed abnormally and often cause drug-resistant epilepsy. Doctors typically treat the condition with surgery, but identifying the lesions from an MRI is an ongoing challenge for clinicians, as FCDs can look normal in scans.
When developing the algorithm, the team, led by University College London (UCL), quantified cortical features from the MRI scans, such as how thick or folded the cortex/brain surface is, and used around 300,000 locations across the brain.
Researchers then trained the algorithm on examples labelled by expert radiologists as either being a healthy brain or having FCD – dependent on their patterns and features.
According to findings published in Brain, the algorithm could detect the FCD in 67 percent of cases in the cohort (538 participants).
Previously, 178 of the participants had been considered MRI negative, meaning that radiologists could not find the abnormality – yet the MELD algorithm could identify the FCD in 63 percent of these cases.
This is “particularly important,” the research team stressed. If doctors can find an abnormality in the brain scan, then surgery to remove it can provide a cure.
“We put an emphasis on creating an AI algorithm that was interpretable and could help doctors make decisions,” said Mathilde Ripart from the UCL Great Ormond Street Institute of Child Health (ICH). “Showing doctors how the MELD algorithm made its predictions was an essential part of that process.”
Around one percent of the world’s population have the serious neurological condition epilepsy, which is characterized by frequent seizures. In the UK, some 600,000 people are affected.
While drug treatments are available for most people with epilepsy. 20-30 percent do not respond to medications.
In children who have had surgery to control their epilepsy, FCD is the most common cause; in adults it is the third most common cause.
Of patients who have epilepsy that has an abnormality in the brain that MRI scans cannot detect, FCD is the most common cause.
Dr Konrad Wagstyl from the UCL Queen Square Institute of Neurology said: “This algorithm could help to find more of these hidden lesions in children and adults with epilepsy, and enable more patients with epilepsy to be considered for brain surgery that could cure epilepsy and improve their cognitive development.”
Wagstyl, a research fellow within the UCL department, added that around 440 children per year could benefit from epilepsy surgery in England.
Dr Hannah Spitzer of German research centre Helmholtz Munich explained that the algorithm automatically learns to detect lesions from thousands of MRI scans of patients. “It can reliably detect lesions of different types, shapes and sizes, and even many of those lesions that were previously missed by radiologists.”
The team remain optimistic about the technology. “We hope that this technology will help to identify epilepsy-causing abnormalities that are currently being missed,” said Dr Sophie Adler from ICH. “Ultimately it could enable more people with epilepsy to have potentially curative brain surgery.”
According to the research team, the study on FCD detection used the largest MRI cohort of FCDs to date, meaning it can detect all types of FCD.
Medics can also run the MELD FCD classifier tool on any patient with a suspicion of having an FCD who is over the age of three and has had an MRI scan.