A unified signal-quality assessment framework for synchronized electro-intravascular encephalography and scalp electroencephalography.

Background

Neurointerventional electrophysiology enables minimally invasive access to near-cortical signals without craniotomy. However, a standardized and interpretable signal-quality evaluation framework for electro-intravascular encephalography (EiEG) is lacking, limiting reproducible reporting and principled comparison with conventional scalp electroencephalography (EEG).

Methods

We proposed a framework quantifying physiological concordance, spectral fidelity, and robustness, and validated it in an ovine EiEG model using a stent-mounted electrode deployed in the superior sagittal sinus over the motor cortex. EiEG and spatially corresponding scalp EEG were synchronously recorded across four repeated 5 min bouts using shared hardware and matched preprocessing.

Results

Using the proposed framework in this preparation, we evaluated physiological concordance, spectral fidelity (signal-to-noise ratio (SNR) index), and robustness in simultaneous EiEG and scalp EEG recordings. Concordance was high (Spearman's ρ>0.94). Under the unified SNR definition, the framework quantified overall linear SNR ratios of 2.04 vs Cz and 1.84 vs C1 over 0.5-40 Hz. Across four repeated 5 min recording bouts, EiEG showed greater consistency than scalp EEG, with lower overall-band across-bout coefficient of variation (CV: 0.8% vs 29.8% for Cz and 6.7% for C1) and lower mean absolute relative deviation from bout 1 (1.5% vs 23.6% and 5.4%, respectively). No acute procedure-related complications were observed.

Conclusions

This framework enables interpretable and reproducible evaluation of EiEG signal quality. When applied to this preparation under synchronized recording, EiEG showed high physiological concordance with scalp EEG together with higher spectral fidelity and greater across-bout consistency.