A University College London research team said its artificial intelligence algorithm could potentially increase the number of drug-resistant epilepsy cases successfully treated with surgery by detecting more cortical abnormalities than humans can detect on their own. (University College London/Konrad Wagstyl).
Epilepsy still has no cure, and while many cases can be managed with anti-seizure drugs, up to 30% of cases are drug-resistant and therefore require more intensive treatment methods.
One such approach, resective surgery, which removes the area of the brain that causes a patient’s seizures, can be very successful. However, these procedures are extremely limited not only by the high risk involved in delving into many areas of the brain, but also by how difficult it can be for doctors to pinpoint the exact cause of epilepsy.
A group of researchers at University College London, who published a study this month in the journal Brain that describes an artificial intelligence algorithm they created to detect the source of seizures in the brain, aims to make that task a little easier.
The researchers focused specifically on cases of focal cortical dysplasia. FCD – areas where brain cells have developed and aggregated abnormally and affect the formation of the cortex – have been reported as a major cause of drug-resistant epilepsy and can be treated with surgery, but are often difficult to detect on standard MRI scans.
They created an artificial intelligence algorithm that was trained to quantify each scan, looking at about 300,000 locations on the brain to measure thickness and detect folds in the cortex, the outermost layer of the brain, and compared them to scans that expert radiologists determined that they have FCD.
After training with half of the scans, the algorithm was tested on the remaining MRIs, representing nearly 550 patients. The AI achieved a sensitivity of just under 60% in recognizing FCD in the scans, a level that increased to 67% when a border zone was added around the lesions in the scans.
In addition, the algorithm was able to accurately detect FCD in 63% of a group of 178 patients whose brain MRIs had previously been declared healthy and free of FCD by radiologists.
The AI tool can be used in any patient who is at least three years old, has undergone an MRI, and is suspected of having FCD.
The researchers said the algorithm could potentially increase the number of drug-resistant epilepsy cases that have been successfully treated with surgery by detecting more cortical abnormalities than can be detected by humans alone. They noted that FCDs are the most common reason for surgery to control epilepsy in children and the third most common in adult patients.