d-EEG-tal

Epilepsy predisposes a person to having recurrent seizures, and it affects almost 1% of the human population. Despite knowledge of this disease and its treatments advances at staggering pace, it is still very difficult to diagnose it properly and swiftly. In fact, the combination of numerous pathologies presenting similar symptoms, together with low sensitivity/specificity of routine exams (brain imaging and electroencephalography (EEG)), leads to a misdiagnosis in the emergency department up to 50%. Moreover, there is ample evidence that delayed treatment is related to a higher risk of relapse, compared to immediate treatment. Thus, a tool able to provide an accurate diagnosis directly after the first seizure is of utmost importance. We started the development of an automated solution, which uses machine learning to extract valuable information from advanced EEG quantitative analysis and clinical information. This tool would provide a diagnosis already at the emergency department, right after the occurrence of the first seizure. If successful, this project could significantly reduce the number of emergency consultations following relapses (while waiting for follow-up exams) and healthcare-related costs, which in turn would drastically improve patients’ quality of life.

The team:

Eric Ménétré

Department of Clinical Neurosciences

Stefano Gallotto

Department of Clinical Neurosciences

Our Current Projects