Fast Ripples are promising electrophysiological biomarkers initially identified at the microscale using tetrodes. Subsequent studies demonstrated their detectability at the macroscale on standard intracerebral contacts, yet macroscopic findings have remained inconsistent regarding their predictive value. FRs are highly focal events whose occurrence and characteristics depend on recording scale. To address this, new deep hybrid electrodes developed with DIXI Medical integrate deployable tetrodes between standard macro-contacts, enabling simultaneous multi-scale recordings in situ.

Fifty-four patients (46 Toulouse, 8 Lyon) were implanted with 173 hybrid electrodes. Recordings were acquired using Blackrock or Neuralynx systems at 2,048 Hz (macro) and >30,000 Hz (micro). FRs were automatically detected during one hour of awake recording per patient using Halyzia software. Safety was assessed through adverse event reporting (ANSM, article R.1123-39 §4) and postoperative MRI. Recording feasibility was evaluated from hourly FR rates at both scales, while biomarker performance was quantified by sensitivity, specificity, and predictive values. All recording contacts were classified based on their localization within the epileptogenic, irritative, or propagation zones.