OptiScales

During presurgical evaluation of epilepsy, digital brain twins allow inferring crucial diagnostic information on tissue excitability based on patient-specific computational models. Current implementation of such twins currently relies on intracerebral EEG signals, which are invasive recordings requiring surgical intervention and long hospitalization time. There is thus a pressing need for a non invasive alternative in patient-specific modeling.

The recent development of 4He optically pumped magnetometers (4He-OPMs) by a French company, Mag4Health, opens the way to unprecedented non invasive recording of brain activity. These sensors benefit from the high spatial discrimination of magnetoencephalography, coupled to an exquisite time resolution, and offer unique additional capacities. They possess a large frequency bandwidth, which can give access to large-scale networks (operating at low frequencies, below 20Hz) and very localized activity (at frequencies above 40Hz, up to 500 Hz). Their spatial capacities can be optimized by bringing sensors close to the scalp in patient specific positions and by recording magnetic fields along several directions. Finally, and importantly, they allow long-term recordings, with the potential to capture very slow fluctuations in excitability as well as the infrequent epileptic seizures.

The objectives of the OptiScales project are two-fold. The first objective is to extract multiscale information from 4He-OPM sensors during epilepsy presurgical evaluation, beyond the capacities of traditional sensors. The second objective is to feed digital twin models with these new signals in order to infer information that is not available directly from the raw signals. Importantly, we will validate signal analysis and model-based inference based on unique invasive recordings in the same patients, thus firmly establishing OPMs as the future of non invasive modelling in epilepsy.

The OptiScales project will gather all the necessary multidisciplinary expertise: multi-scale signal acquisition, signal analysis, signal modelling, clinical interpretation, in collaboration with the company developing the technology. It will build and strengthen links in France around OPMs. It will have an impact on signal analysis, computational neuroscience and clinical practice. Finally, it will make France a key player in the booming OPM technology.