Time-frequency signatures evoked by single-pulse deep brain stimulation to the subcallosal cingulate

Ezra E. Smith, Ki Sueng Choi, Ashan Veerakumar, Mosadoluwa Obatusin, Bryan Lad Howell, Andrew H. Smith, Vineet Tiruvadi, Andrea Crowell, Patricio Riva‐Posse, Sankaraleengam Alagapan, Christopher J. Rozell, Helen S. Mayberg, Allison C. WatersFrontiers in Human Neuroscience 2022 · DOI

TL;DR

This study explores how the precise location of deep brain stimulation (DBS) in the subcallosal cingulate (SCC) affects brain activity, measured by EEG. They found that subtle millimeter-scale changes in stimulation site significantly alter brain's electrical responses, particularly in theta frequency phase clustering, suggesting a new way to optimize DBS targeting for treatment-resistant depression.

What Problem Does This Paper Address?

Deep brain stimulation (DBS) is a promising treatment for conditions like treatment-resistant depression (TRD), but its effectiveness hinges on precisely targeting specific brain regions and white matter pathways. The subcallosal cingulate (SCC) is a key target, but its complex structure makes optimal electrode placement challenging. Current methods for confirming target engagement are limited. This paper addresses the need for more refined methods to assess the impact of DBS stimulation location on brain activity, moving beyond simple event-related potentials (ERPs) to explore time-frequency dynamics. Understanding these dynamics could lead to more personalized and effective DBS treatments.

Methods

The researchers used electroencephalography (EEG) to record brain activity from eight patients undergoing DBS for TRD. The core manipulation involved delivering single electrical pulses from different contacts on the implanted DBS leads within the SCC target region. This allowed them to investigate how varying the stimulation location (dorsal vs. ventral contacts) affected the brain's response.

Key techniques included:

  • Time-Frequency Analysis: Instead of just looking at the overall timing of brain responses (ERPs), they used Morlet wavelets to analyze how spectral power and inter-trial phase clustering (ITPC) changed over time and across different frequencies. This provides a richer picture of neural dynamics.
  • Stimulation-Evoked Potential (SEP) Analysis: They analyzed the evoked potential (SCC-EP) that occurs in response to each stimulation pulse, focusing on a specific time window (50-150 ms) known to be relevant.
  • White Matter Density Analysis: They correlated the observed EEG signals with the amount of white matter tissue activated by the DBS contacts, using tractography and CT imaging to estimate the volume of tissue activated (VTA).

Datasets

The study utilized data from eight patients with treatment-resistant depression who had undergone subcallosal cingulate (SCC) deep brain stimulation (DBS) surgery. This is a patient cohort, not a publicly available dataset.

Key Insights and Ideas

  • Millimeter-scale precision matters: Even tiny shifts in DBS electrode placement within the SCC target region lead to measurable differences in cortical electrical activity.
  • Theta phase clustering as a marker: Enhanced phase clustering in the theta frequency band (4-7 Hz) consistently appeared in response to SCC stimulation and was strongly linked to the magnitude of the evoked potential, especially with left-sided stimulation. This suggests theta phase synchrony is a key component of the brain's response.
  • Dorsal stimulation yields stronger effects: Stimulation targeting more dorsal parts of the SCC showed stronger theta phase clustering compared to ventral stimulation. This hints at differential engagement of white matter pathways.
  • Shared generative mechanisms: The findings suggest that the mechanisms underlying stimulation-evoked potentials might be similar to those involved in responses to sensory stimuli or during cognitive tasks, particularly concerning phase clustering.
  • Potential for objective targeting: Time-frequency features like theta phase clustering could serve as objective markers to confirm optimal DBS target engagement, moving beyond subjective patient reports or less precise imaging.

Key Results

  • SCC-EP Magnitude: Stimulation location significantly affected the magnitude of the SCC-EP, with a linear decrease observed as stimulation moved from dorsal to ventral contacts.
  • Theta ITPC: Theta ITPC (4-7 Hz) was significantly affected by stimulation location, with stronger phase clustering observed with dorsal stimulation. An interaction between hemisphere and contact location was also significant.
  • Correlation between ITPC and SCC-EP: Theta ITPC was significantly correlated with SCC-EP magnitude, particularly for left SCC stimulation (Spearman's r = -0.57, p < 0.001).
  • White Matter Correlation: A weak but significant association was found between activated white matter volume and theta ITPC (R² = 0.13, p < 0.01).
  • Delta Power: An increase in evoked delta power (2-4 Hz) was observed coincident with SCC-EP but was less consistent across participants.

Limitations and Future Work

  • Non-randomized Stimulation: The order of stimulation conditions was not randomized, which could introduce confounding effects of stimulation sequence on neural responses. Future work should address this.
  • Stimulator Artifacts: While advanced preprocessing techniques were used, stimulator artifacts remain a significant challenge in DBS-EEG research. The authors acknowledge that their artifact mitigation might have attenuated some neurogenic activity.
  • Right SCC Stimulation: The correlation between ITPC and EP was not significant for right SCC stimulation, which might be related to a weaker overall EP response observed in the right hemisphere.
  • Future Directions: The authors suggest that future research should explore symmetric biphasic pulses to minimize artifacts, investigate the influence of stimulation sequence, and further examine how these time-frequency features can differentiate specific white matter bundles for therapeutic targeting. They are optimistic that these methods can help optimize DBS placement and personalize treatment for conditions like depression.

Summary updated

Abstract

Precision targeting of specific white matter bundles that traverse the subcallosal cingulate (SCC) has been linked to efficacy of deep brain stimulation (DBS) for treatment resistant depression (TRD). Methods to confirm optimal target engagement in this heterogenous region are now critical to establish an objective treatment protocol. As yet unexamined are the time-frequency features of the SCC evoked potential (SCC-EP), including spectral power and phase-clustering. We examined these spectral features-evoked power and phase clustering-in a sample of TRD patients (<i>n</i> = 8) with implanted SCC stimulators. Electroencephalogram (EEG) was recorded during wakeful rest. Location of electrical stimulation in the SCC target region was the experimental manipulation. EEG was analyzed at the surface level with an average reference for a cluster of frontal sensors and at a time window identified by prior study (50-150 ms). Morlet wavelets generated indices of evoked power and inter-trial phase clustering. Enhanced phase clustering at theta frequency (4-7 Hz) was observed in every subject and was significantly correlated with SCC-EP magnitude, but only during left SCC stimulation. Stimulation to dorsal SCC evinced stronger phase clustering than ventral SCC. There was a weak correlation between phase clustering and white matter density. An increase in evoked delta power (2-4 Hz) was also coincident with SCC-EP, but was less consistent across participants. DBS evoked time-frequency features index mm-scale changes to the location of stimulation in the SCC target region and correlate with structural characteristics implicated in treatment optimization. Results also imply a shared generative mechanism (inter-trial phase clustering) between evoked potentials evinced by electrical stimulation and evoked potentials evinced by auditory/visual stimuli and behavioral tasks. Understanding how current injection impacts downstream cortical activity is essential to building new technologies that adapt treatment parameters to individual differences in neurophysiology.
BibTeX
@inproceedings{smith2022time,
  author = {Ezra E. Smith and Ki Sueng Choi and Ashan Veerakumar and Mosadoluwa Obatusin and Bryan Lad Howell and Andrew H. Smith and Vineet Tiruvadi and Andrea Crowell and Patricio Riva‐Posse and Sankaraleengam Alagapan and Christopher J. Rozell and Helen S. Mayberg and Allison C. Waters},
  title = {{Time-frequency signatures evoked by single-pulse deep brain stimulation to the subcallosal cingulate}},
  booktitle = {Frontiers in Human Neuroscience},
  year = {2022},
  doi = {10.3389/fnhum.2022.939258},
}
Figure