Back

Chronic sensing during subcallosal cingulate deep brain stimulation captures ketamine-associated changes in major depressive disorder

Sumarac, S.; Lawson, T.; Favi Bocca, L.; Rabin, J. S.; Giacobbe, P.; Nestor, S.; Hamani, C.; Lipsman, N.; Milosevic, L.; Davidson, B.

2026-09-11 psychiatry and clinical psychology
10.64898/2026.09.08.26362080 medRxiv
Show abstract

Ketamine and subcallosal cingulate (SCC) deep brain stimulation (DBS) have each shown antidepressant effects in treatment-resistant depression, yet how ketamine modulates SCC physiology during chronic stimulation remains poorly understood. We analyzed 5,907 hours of bilateral SCC sensing, 485 participant-triggered recordings across 193 days, 47 ketamine dosing episodes, and an in-clinic session. Across naturalistic and in-clinic recordings, ketamine was associated with a convergent bilateral shift toward 21-40 Hz fast-band activity. Post-dose snapshots showed lower aperiodic-adjusted 5-20 Hz power and higher 21-40 Hz power, while 23.44 Hz activity increased bilaterally after administration. Longitudinal 23.44 Hz tracking revealed larger bilateral increases around anxious and sad events, beginning before self-report and peaking approximately 15 minutes later, smaller decreases after neutral and happy events, and smaller lateralized changes around dosing, despite limited spectral separation between affective states at event onset. These findings provide direct human evidence that ketamine exposure and affective states can engage the same SCC frequency range with distinct spectral and temporal expressions. This positions SCC fast-band activity as a context-dependent signal whose interpretation may depend on its magnitude, laterality, and time course, motivating adaptive DBS strategies that decode evolving SCC dynamics rather than rely on simple band-power thresholds.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.