Multi-centre Epilepsy Lesion Detection project: bringing AI into epilepsy imaging

Focal lesions are a common cause of drug-resistant epilepsy that can be challenging to identify on MRI. However, when detected, they can often be treated effectively through surgical resection. The Multi-centre Epilepsy Lesion Detection (MELD) Project is an international collaboration dedicated to improving the detection of Focal Cortical Dysplasias and other lesions in patients with drug-resistant epilepsy. In this talk, I will present our work on developing and training AI tools to detect 1) focal cortical dysplasia, 2) hippocampal sclerosis and finally multiple epilepsy pathologies using a large multicentre MRI cohorts of patients with focal lesions. I will discuss how our pipelines are designed to provide interpretable reports, capturing lesion location but also model confidence and describe feature abnormalities and how salient they are to the algorithms decision.  Finally, I will share updates on how our open source MELD tools are being integrated into presurgical planning for epilepsy at Great Ormond Street Hospital (GOSH) and other centers around the world.