Scientists have developed an AI-powered tool that detects 64% of brain abnormalities linked to epilepsy that human radiologists miss, according to a new study published in Jama Neurology.
鈥淭his tool could drastically change the diagnostics and care for 2 500 patients in Finland and four million worldwide,鈥 says Docent Yawu Liu, who participated in the study from the 91天美.
Led by a team at King鈥檚 College London and University College London (UCL), the project also involved the Epilepsy Biomarker Study led by Professor Reetta K盲lvi盲inen at the 91天美 and Kuopio University Hospital Epilepsy Centre. In Finland, the study was funded by the Saastamoinen Foundation and the Research Council of Finland.
The AI tool called MELD Graph significantly improves the detection of focal cortical dysplasias (FCDs), a leading cause of epilepsy. 鈥淭his will speed up diagnosis times and get patients the surgical treatment they need quicker,鈥 K盲lvi盲inen says.
The researchers estimate that for example in the UK, the AI tool could reduce costs to the NHS by up to 拢55,000 per patient.
1 in 100 people are affected by epilepsy, and 1 in 5 people with epilepsy have seizures caused by a structural abnormality (鈥渓esion鈥) in the brain. FCDs are a common structural cause of epilepsy and in people with this type of epilepsy, seizures can鈥檛 usually be controlled with medications. Surgery to remove the lesion can be an effective and safe way to stop the seizures. However, the challenge is that FCDs can be subtle and difficult to see with the human eye. Up to half of these lesions are missed by radiologists. Delays to diagnosis and surgery mean more seizures, more emergency visits, and more disruption to school, work and home life.
In the study, the researchers pooled MRI data from 1185 participants 鈥 including 703 people with FCD and 482 controls 鈥 from 23 epilepsy centres around the world in the Multicentre Epilepsy Lesion Detection project (MELD). They then trained the artificial intelligence tool on the scans to detect these subtle brain abnormalities that might otherwise go undetected.
According to K盲lvi盲inen, the use of an AI-powered tool such as MELD Graph would undoubtedly also benefit the Finnish healthcare system by supporting neuroradiologists with their decisions, speeding time to treatment for patients and relieving them of unnecessary and costly tests and procedures.
While the tool is not yet clinically available, the research team have released it as an open-source software. They are running workshops to train clinicians and researchers around the world, including within the European Reference Network for rare and complex epilepsies, EpiCARE ERN.
For further information, please contact:
Reetta K盲lvi盲inen, Professor of Neurology, reetta.kalviainen@uef.fi
Research article:
Ripart M et al., 2025. . JAMA Neurology, Feb 24. doi:10.1001/jamaneurol.2024.5406
