A team of international researchers has developed an artificial intelligence (AI) algorithm that can detect subtle brain abnormalities which cause epileptic seizures.1
The researchers led the Multicentre Epilepsy Lesion Detection project (MELD), which used over 1,000 patient MRI scans from 22 global epilepsy centres to develop the algorithm.
The algorithm provides reports of where brain abnormalities (or lesions) are. It can reliably detect lesions of different types, shapes and sizes, and even lesions that were previously missed by radiologists.
This is particularly useful 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. It is typically treated with surgery, however identifying the lesions from an MRI is an ongoing challenge for clinicians, as MRI scans in FCDs can look normal.
Epilepsy is a serious neurological condition that affects the brain and causes frequent seizures. It is very common in people with learning disabilities, with about one in three with a mild to moderate learning disability having the condition.2
According to the Epilepsy Society, generally, having a learning disability does not cause epilepsy, and having epilepsy does not cause a learning disability. Some people may have epilepsy and learning disabilities, and both may be caused by the same underlying problem in the way their brain works. That being said, the more severe the learning disability, the more likely that the person will also have epilepsy.2
While drug treatments are available to the majority of people with epilepsy, around 20-30% do not respond to medication. The authors of the research therefore hope that the new algorithm will enable more patients to be considered for brain surgery that could cure the epilepsy and improve their cognitive development.
The team, led by researchers from University College London (UCL), quantified cortical features from the MRI scans, such as how thick or folded the cortex/brain surface was, and used around 300,000 locations across the brain.
Expert radiologists then identified the features as either a healthy brain or as having FCD and taught the algorithm to categorise the features into two groups. Study results showed the algorithm was able to detect the FCD in 67% of cases in the cohort (538 participants).
Previously, radiologists were unable to find the abnormality in 178 of the participants, yet the MELD algorithm was able to identify the FCD in 63% of these cases.
Co-senior author, Dr Konrad Wagstyl (UCL Queen Square Institute of Neurology) therefore hopes the algorithm could help to find more of these “hidden lesions” in adults and children with epilepsy.
He said: “[This could] enable more patients with epilepsy to be considered for brain surgery that could cure the epilepsy and improve their cognitive development. Roughly 440 children per year could benefit from epilepsy surgery in England.”
