🧠 We are #hiring! Postdoc @stanfordmedicine.bsky.social: help us develop closed-loop EEG platforms to improve TMS treatment for depression! Coding and neuroscience / engineering backgrounds required. Join our team! precisionneuro.stanford.edu email: ckeller1@stanford.edu
When we use TMS to treat depression, we think it alters neural activity in a deep brain structure known as the subgenual anterior cingulate (sgACC). But this is hard to test directly, until our preprint with @coreykeller.bsky.social, @neuro-engineer.bsky.social, Nick Trapp, and Aaron Boes (UIowa) 1/
🧠 NEW PAPER: How do we capture excitability noninvasively in mood and emotion networks in the human brain? Here, we used real-time optimization to improve these measures!
@sparmi.bsky.social @stanfordpntlab.bsky.social @clinicalneuroph.bsky.social doi.org/10.1016/j.clinph.2025.02.261 1/7
And to the rest of the amazing team and resources! 7/7
@coreykeller.bsky.social
@juhagogulski.bsky.social
@jessicamross8.bsky.social
@stanfordpntlab.bsky.social
@stanfordmedicine.bsky.social
Manjima Sarkar, Jade Truong, Lily Forman!!
Big shout out to my co-first author 👏
@chrisclineneuro.bsky.social
Interested in learning more? Read the full paper in Clinical Neurophysiology! doi.org/10.1016/j.cl... 5/7
Key findings: Optimization reduced artifacts by 63% and increased early local TMS-evoked potentials (EL-TEPs), a measure of prefrontal excitability, by 75%! #EEG #TMS #TMSEEG 4/7
We introduce a novel method for optimizing TMS parameters in the dlPFC. Based on EEG responses, this closed-loop procedure optimizes TMS coil angle, location, and intensity in real time. 3/7
Why is this important? The dlPFC is a depression treatment target, but we need clinic-ready ways to measure stimulation effects. TMS-EEG can help, but artifacts obscure responses. 2/7
🧠 NEW PAPER: How do we capture excitability noninvasively in mood and emotion networks in the human brain? Here, we used real-time optimization to improve these measures!
@sparmi.bsky.social @stanfordpntlab.bsky.social @clinicalneuroph.bsky.social doi.org/10.1016/j.clinph.2025.02.261 1/7
This builds on influential work by
@foxmdphd.bsky.social
@shansiddiqi.bsky.social
@desmondoathes.bsky.social and many others. It would not have been possible without the support of UIowa, Stanford, and generous funding from the NIH/NIMH.
reposted from @esolomon.bsky.social
It’s important to be clear that we only had two neurosurgical patients to test these effects, so these findings need replication. But these data so far seem to align with a major hypothesis in the field: DLPFC TMS specifically alters population-level neural activity in the sgACC. 5/
We were surprised to find that, despite anticorrelated HFA signals, TMS seemed to increase phase-locking in lower frequencies (alpha and theta) between sgACC-DLPFC. Could this be a mechanism by which the DLPFC influences sgACC activity? 4/
HFA signal in the sgACC was inversely correlated with HFA signal in the DLPFC, meaning that as DLPFC activity increased following TMS, sgACC activity decreased. This mirrors what we’ve known from fMRI for a long time, but now shown with direct in-vivo measures of neural activity. 3/
Stimulation to the DLPFC may tamp down an overactive sgACC and relieve symptoms of depression. By stimulating the DLPFC with TMS in two neurosurgical patients with electrodes implanted within the sgACC, we found a reduction in high-frequency neural activity – a correlate of population spiking. 2/
When we use TMS to treat depression, we think it alters neural activity in a deep brain structure known as the subgenual anterior cingulate (sgACC). But this is hard to test directly, until our preprint with @coreykeller.bsky.social, @neuro-engineer.bsky.social, Nick Trapp, and Aaron Boes (UIowa) 1/
🚨 New paper alert! Congratulations to the team on getting their paper accepted to @clinicalneuroph.bsky.social 🥳👏
@sparmi.bsky.social @chrisclineneuro.bsky.social @juhagogulski.bsky.social @jessicamross8.bsky.social @coreykeller.bsky.social
Read more below 👇 @stanfordbrain.bsky.social
Preprint alert! Check out our latest work in @stanfordpntlab.bsky.social, with @coreykeller.bsky.social @neuro-engineer.bsky.social @esolomon.bsky.social Nick Trapp Aaron Boes (UIowa) 👏
Congrats to our 2025 Neuroscience:Translate awardees! Their projects support brain imaging and stimulation technologies to improve depression treatment, broaden access to TMS therapy, and advance methods for imaging inflammation in the brain.
Learn more: neuroscience.stanford.edu/news/brain-i...
🚀 Exciting week at #BrainStimConf in Japan 🇯🇵!
@coreykeller.bsky.social @juhagogulski.bsky.social & @chrisclineneuro.bsky.social are presenting on #TMSEEG, plasticity & precision neuromodulation 🧠⚡
See the full lineup below! 📸✨ @stanfordmedicine.bsky.social #neuroscience
🧠 We are #hiring! Postdoc @stanfordmedicine.bsky.social: help us develop closed-loop EEG platforms to improve TMS treatment for depression! Coding and neuroscience / engineering backgrounds required. Join our team! precisionneuro.stanford.edu email: ckeller1@stanford.edu
and @chrisclineneuro.bsky.social!
1/10 🧠 We need better ways to measure brain excitability in psychiatric disorders. The prefrontal cortex—key target for brain stimulation treatments—is particularly hard to assess. Our new studies tackle this with systematic mapping! bit.ly/4cjHknX & bit.ly/3PW8RCx
9/9 @juhagogulski.bsky.social @coreykeller.bsky.social @stanfordpntlab.bsky.social @jessicamross8.bsky.social @sparmi.bsky.social @stanfordmedicine.bsky.social @stanfordpress.bsky.social @nihr.bsky.social Truong J., Sarkar M.
8/9 🔑 Key takeaways:
• Careful target selection boosts early TMS-evoked signals in dlPFC
• Most dlPFC areas can produce reliable TMS-evoked signals with optimized analysis
This work advances early TMS-evoked EEG signal as a potential biomarker for depression! #depression
7/9 ⚡️ Surprisingly, we found that as few as 25 TMS pulses could produce reliable responses from the medial dlPFC target! This is much lower than previously thought and could allow for faster measurements.
6/9 💪 We found that reliability was improved by using:
• Peak-to-peak amplitude
• Later time windows (20-60 ms or 30-60ms)
• Sensor-space (vs. source-space) analysis
ROI size didn't impact reliability much.
5/9 🔍 We then examined how reliable these early TMS-evoked responses are. The most medial dlPFC target was most reliable, while the most anterior was least reliable. But with optimized analysis, most targets could produce reliable signals. #reliability
4/9 📈 The "best" dlPFC target produced responses over 2x larger than standard clinical targets! This suggests we may be able to boost the early TMS-EEG signal by carefully selecting the stimulation location. #optimization
3/9 📊 We found posterior & medial targets produced larger responses with less muscle artifact compared to anterior & lateral targets. This suggests some areas of the dlPFC may be better suited for measuring brain excitability with TMS.