Spectral separation of evoked and spontaneous cortical activity, Part 2: Somatosensory high frequency oscillations

G. Fischer,J. Haueisen,D. Baumgarten,M. Kofler
DOI: https://doi.org/10.1016/j.bspc.2024.106456
IF: 5.1
2024-05-24
Biomedical Signal Processing and Control
Abstract:N -Interval Fourier Analysis ( N -FTA) allows for simultaneous spectral assessment of evoked and spontaneous activity in the frequency domain. We applied this method to signals following peripheral electrical nerve stimulation and performed analysis of cortical somatosensory evoked potentials within the 400 to 750 Hz band. For median nerve stimulation, data from eleven volunteers were analyzed. For tibial nerve stimulation, three subjects were investigated. For both stimulation sites, evoked high frequency oscillations (HFOs) components were identified. Furthermore, two kinds of background HFO activity were detected in sham stimulation trials. Spectral component models were applied for quantifying signal properties. Evoked spectral components reflected HFOs being time-locked to the stimulus. The detected spectral components were distributed over the entire investigated spectral band. Their spectral amplitude was close to the limit of the resolution of N -FTA. The experimentally observed spectral amplitude were in quantitative agreement with a model using a Morlet morphology. Within the HFO band, a flat noise floor was observed. Spontaneous physiological background activity contributes significantly to the spectral amplitude. This random activity is the dominant source of interference when extracting evoked HFOs. Within the HFO band, narrow spectral peaks in background activity were detected – both for real and sham stimulation. In the data sampled at 9.6 kHz, such peaks were observed in all recordings. For the 5.0 kHz sampling rate, these peaks were visible in about half of the recordings, and their amplitude was reduced. Based on a mathematical model, these peaks may be generated by organized spontaneous HFO activity producing a stable background wave.
engineering, biomedical
What problem does this paper attempt to address?