This algorithmic program was used in the Multi Center Epilepsy Lesion Detection (MELD) project. It reported the location of anomalies in drug-resistant focal cortical dysplasia (FCD). FCD is a major cause of epilepsy. The program is based on the analysis of MRI’s of over 1000 patients from 22 international epilepsy centres.
Improperly having FCD in the brain often causes drug-resistant epilepsy. Usually, the procedure of surgery is adopted for its treatment. However, the detection of lesions on MRI remains an ongoing problem for doctors, as an MRI scan may appear normal for FCD.
The scientists tried to estimate the cartilaginous properties from MRI scans by examining more than three lakh locations throughout the brain. In this, it was to find out how thick or bent the surface of the cortex or brain is.
Based on this, radiologists decide whether the brain is healthy or has FCD. It tells the pattern of the brain. In general, this algorithm was successful in detecting FCD by 67 percent of the 538 participants included in the study, according to this research findings published in the journal Brain.
Earlier radiologists had failed to detect abnormalities in the brain of 178 patients based on MRI. Whereas with the MELD algorithm they were able to detect FCDs in 63 percent of them.
This is important because if doctors can identify anomalies from brain scans, it will be easier to perform precise surgery to remove them.
Mathilde Ripart, a researcher at the UCL Great Ormond Street Institute of Child Health and the first co-author of this research, said that in developing the AI system, they focused on how their analysis would help doctors make decisions. An important step in this process was to demonstrate to doctors how the MELD algorithm is capable of predicting disease.
Dr Konard Vagstyle, a researcher at the UCL Queen Square Institute of Neurology and co-senior author of the study, said the algorithm would make it easier to identify latent lesions in the brains of epileptic patients. This will increase the number of patients who can benefit from surgery. In addition, cognitive function can also be improved in patients.