Pathological High-frequency oscillations (HFOs) are short-duration (a few tens of ms), low-amplitude signals occurring in electroencephalographic (EEG) recordings in a specific frequency band (250-600 Hz). Their clinical value has been reported in a number of studies showing that i) HFOs can be observed with intracranial macro-electrodes in patients with refractory focal epilepsy awaiting surgery, ii) the occurrence rate of HFOs is higher within the seizure onset zone in most patients and iii) there is a correlation between the resection of brain sites generating HFOs and favorable surgical outcome.
The objective of the NEURO-SENSE project is to increase the diagnostic value and routine use of HFOs as a reliable marker of the epileptogenic zones (EZ).
To reach this objective, we will solve a number of scientific and technological issues. Firstly, the relationship between the epileptic tissue characterized by an impaired balance between excitation/inhibition processes, and the spectral features of HFOs as observed on multi-contact intracranial electrodes is not straightforward. There is still a lack of understanding on the specific information that is carried by HFOs. Secondly, the properties of recording electrodes (technology, material, form factor, position w.r.t. neuronal sources, electrode-tissue interface) dramatically impact the observability of HFOs that often mix with background interictal activity. Thirdly, automatic detection of HFOs, which is recognized as a difficult problem due to poor signal-to-noise ratio, wide morphological/spectral diversity and signal contamination by artifacts must still be improved. And fourthly, the design of human clinical electrodes, classically used during pre-surgical evaluation of patients candidate to surgery, can likely be revisited in order to optimize the recording of HFOs.
To address these issues, our working hypothesis is that novel extracellular electrodes can be designed in order to optimize the recording and the detection of pathological HFOs in brain signals. The first specific aim is to develop new “hybrid” computational models integrating i) pathophysiological neuronal mechanisms at the origin of field-recorded HFOs and ii) biophysical information regarding the electrode form factor and electrode-tissue interface. Based on model-guided predictions regarding the observability of HFOs, the second specific aim is to build and test novel small-scale (10-100 µm) field electrodes optimized for HFO recording in vivo (animal model of epilepsy). Our recordings will offer the unique opportunity to address the third specific aim: propose high-performance detection algorithms able to specifically extract pathological HFOs from local field potentials. Finally, the last specific aim is to design and test, in a simulation environment, virtual larger-scale (1-2 mm) intracerebral clinical electrodes optimized for HFO recording in the human brain.
The NEURO-SENSE project addresses the ANR 2018 work program challenge titled “Life, health and well-being” and theme 9 “Healthcare technologies”. Technological research performed in the project will lead to improved detection of pathological high-frequency oscillations. Based on a unique combination of computational modeling, signal processing and microfabrication of implantable probes, NEURO-SENSE will deliver a new generation of biosensors that significantly improve the extraction of relevant markers of epilepsy in brain signals. Technological developments will also include novel neural implants as biocompatibility and long-term recording under chronic conditions will also be improved.
Overall, NEURO-SENSE will contribute to future generation of medical devices aimed at improving diagnosis in severe drug-resistant epilepsies which are recognized as a major health problem.
Monsieur Fabrice Wendling (LABORATOIRE TRAITEMENT DU SIGNAL ET DE L'IMAGE)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
LTSI LABORATOIRE TRAITEMENT DU SIGNAL ET DE L'IMAGE
INS Institut de Neurosciences des Systèmes
Mines Saint Etienne / CMP / Bioélectronique Ecole Nationale Supérieure des Mines Saint-Etienne – Ecole de l’Institut Mines Télécom.
Help of the ANR 637,664 euros
Beginning and duration of the scientific project: March 2019 - 48 Months