Authors :
Presenting Author: Nicholas Reyes, BS – University of Alabama at Birmingham
Helen Brinyark, BS – University of Alabama at Birmingham
Caila Coyne, MEng – University of Alabama at Birmingham
Benjamin Cox, MD – University of Alabama at Birmingham
Arie Nakhmani, PhD – University of Alabama at Birmingham
Rachel Smith, PhD, MS, BS – University of Alabama at Birmingham
Rationale:
Surgical resection of the seizure onset zone (SOZ) is currently the most effective treatment method for drug-resistant epilepsy. When implanted with intracranial EEG electrodes to localize the SOZ, the clinical team may use single-pulse electrical stimulation to better identify epileptogenic regions.
Our group has found that dynamical network models constructed from stimulation-evoked potentials enable identification of “resonant frequencies”, specific frequencies that produces large amplitude oscillatory responses in the EEG that may propagate seizure activity (Smith et al., 2022). These resonant frequencies are identified as a sharp peak in the magnitude versus frequency response curve, called a Bode plot. We hypothesize that neural resonance is likely anatomically specific, but this has not been tested. Identifying physiological resonant frequencies across the human brain can help identify when pathological resonance may be driving seizure initiation.
Methods:
We included 5 patients who underwent intracranial monitoring for whom we had MRI and CT scans available at UAB. We constructed Bode plots from each patient’s stimulation responses to measure the local response magnitude of the brain to a range of input stimulation frequencies for each electrode pair. Then we used RAVE 2.0 to localize each patient’s electrodes to individual anatomical regions using the N27 brain atlas. We identified peaks within the Bode plot which would represent resonant frequencies from 5 Hz to 30 Hz. To measure the power of a peak, we used a custom equation using the peak’s maximum response, the prominence of the peak, and the width of the peak. We then normalized the values to a range of values between 1 and 10, with 10 representing very sharp, prominent peak with small width. We picked the top three strongest peaks, rounded them to the closest whole number frequency and identified other peaks that rounded to the frequency.
Results:
We found 3 out of 5 patients had stimulations and responses on both hemispheres of the temporal middle gyrus. Of these 3 patients, all three had 7 Hz as one of their top three strongest peaks and two had 5 Hz and 8 Hz as their strongest peaks when individually analyzed. The top three most prominent frequencies from all the patients were 5, 6, and 7 Hz respectively. The average power of 7 Hz for each patient typically was lower than other strong peaks but consistently found to be one of the strongest peaks. The variance for the peak power at 5 Hz, 6Hz, and 7Hz was 1.36 Hz, 1.59 Hz, and 1.16 Hz, respectively.
Conclusions:
These findings suggest that there may be anatomically specific resonant frequencies across patients. Identification of anatomically specific resonant frequencies can help distinguish physiological and pathological resonant frequencies which could have implications for identifying frequencies associated with neurological disorders such as epilepsy. Further investigation could create a baseline map of physiological resonance from which we could identify biomarkers of the SOZ.
Funding:
This work was funded by AES Junior Investigator Award 1042632 and CURE Epilepsy Taking Flight Award 1061181.