Lukas Goede

What might the future of deep brain stimulation look like?

Not just continuous stimulation, but decoding symptoms in real time and directing stimulation to the right brain circuits.

Fascinating perspective article by @andreashorn.org & @julianneumann.bsky.social in @natrevneurol.nature.com:

Is there a shared brain circuit target for tremor that transcends diagnosis and DBS site? New work by @lukasgoede.bsky.social @andreashorn.org @braincircuits.bsky.social says YES

And of course, institutions like Brigham and Women’s Hospital, Boston, and Charité - Universitätsmedizin Berlin, providing the environment that made this work possible.

Projects like this are only possible thanks to strong research support.

Grateful for funding from the Thiemann Parkinson Foundation, @dfg.de, and many others who made this work possible - not just for the project, but personally for me as well.

Special thanks to @andreashorn.org and @foxmdphd.bsky.social for their mentorship at the @braincircuits.bsky.social, BWH.

This was a long-term project developed under @andreashorn.org's guidance, and I couldn’t be more grateful that it marks the conclusion of my time in Boston.

Thank you to patients, collaborators and coauthors who made this project possible!

To mention just a few colleagues who are here on BlueSky:
@patriciazvarova.bsky.social @bahnebahners.bsky.social @emiddlebrooksmd.bsky.social @jjoutsa.bsky.social @foxmdphd.bsky.social @andreashorn.org

@netstim.org

Convergent mapping of a tremor treatment network - Nature Communications Goede et al. combined multiple modalities to define a common tremor network across disorders. This finding may help optimize deep brain stimulation and guide future noninvasive therapies.

This was a huge team effort - combining datasets across disorders, continents, and methods.

We hope this work helps advance neuromodulation toward guided, symptom-specific treatment.

Full paper: www.nature.com/articles/s41...

We also provide a literature overview of a key structure consistently associated with tremor improvement:
👉 the dentato-rubro-thalamic tract (DRT).

The best DBS outcomes align closely with its trajectory - regardless of target or disorder.

We added extensive supplementary material, including a review on:
👉 What influences outcomes after DBS?

Of course, there are limitations.

We combined diverse datasets and methods - adding variability.
But this heterogeneity was deliberate, to build a more robust and generalizable network.

And while correlational analysis, the convergence of lesions, DBS, and EMG-fMRI supports causal insight.

So what do we take away from all this?

✅ Different forms of tremor share a common brain network
✅ DBS likely works by modulating this circuit
✅ STN, VIM, GPi may all be access points to the same network
✅ This opens the door to symptom-specific neuromodulation - both invasive and noninvasive

And yes - indeed.

Connectivity between GPi-DBS electrodes and the convergent tremor map explained significant variance in clinical outcomes.

We then integrated all maps - lesions, atrophy, fMRI, and DBS - into a multimodal tremor network map.

The key test:
Could this convergent map explain tremor outcomes in a new, independent cohort: Parkinson's disease patients with GPi-DBS?

To power these analyses, we created a new, high-resolution normative connectome:
🧠 Based on resting-state fMRI from 1,087 healthy subjects (HCP), 2 mm isotropic voxels

We also replicated results using a Parkinson’s disease-specific connectome from the PPMI cohort. Findings held up across datasets.

Building on these data, we analyzed two different DBS patient cohorts:

• Parkinson’s disease patients with subthalamic DBS
• Essential tremor patients with thalamic DBS

Outcome maps from each group could explain outcomes in the other.

Disorder-independent. Target-independent.

Another key step was to collaborate with Rick Helmich.

We incorporated a map from the well-known dimmer-switch model of tremor, based on EMG-fMRI:
🔗 academic.oup.com/brain/articl...

Once again, the same core regions emerged:
- Motor cortex
- Motor cerebellum

Three paths. One destination.

Mapping Essential Tremor to a Common Brain Network Using Functional Connectivity Analysis | Neurology Background and ObjectivesThe cerebello-thalamo-cortical circuit plays a critical role in essential tremor (ET). However, abnormalities have been reported in multiple brain regions outside this circuit...

Next, we included an atrophy network map from essential tremor patients, based on a recent paper by Ellen Younger et al.:
🔗 doi.org/10.1212/WNL....

This atrophy-based network pointed to cerebellar regions, adding another independent layer of support.

We found: the stronger the connection between a DBS site and this lesion-derived network, the better the tremor relief - even across different disorders.

Identifying therapeutic targets from spontaneous beneficial brain lesions Brain damage can occasionally result in paradoxical functional benefit, which could help identify therapeutic targets for neuromodulation. However, these beneficial lesions are rare and lesions in mu...

First, we revisited stroke lesions that alleviated tremor.
As shown by @jjoutsa.bsky.social‬, these lesions mapped to a distinct functional network:

🔗 doi.org/10.1002/ana....

We asked a simple question:

Does a shared tremor treatment network exist across diseases and deep brain stimulation (DBS) targets?

To find out, we combined four independent modalities:
🧠 Lesions
🧠 Atrophy
🧠 EMG-fMRI
🧠 DBS outcomes

Tremor is a common symptom across many neurological disorders.

But is there a shared tremor circuit across disorders - one that could guide treatment, regardless of diagnosis?

We think: yes.

Our study is out now in @natcomms.nature.com:

🔗 www.nature.com/articles/s41...

🧵 A thread.

Patient Selection in Deep Brain Stimulation: A Role for Transcranial Direct Current Stimulation to Enhance the Levodopa Challenge? Dopaminergic medication and deep brain stimulation (DBS) improve motor symptoms in Parkinson's disease (PD), but levodopa response alone may not predict DBS outcomes. We retrospectively analyzed 19 P....

Big thanks to @patriciazvarova.bsky.social and @bahnebahners.bsky.social for their wonderful contributions, and to @andreashorn.org for his mentorship on this project!

Read the full study here:

onlinelibrary.wiley.com/doi/10.1002/...

Of course: this is a small sample and an exploratory analysis.

But this cohort is rare: patients who received all three interventions: tDCS, levodopa challenge, and DBS.

That makes these findings hypothesis-generating and worth building on.

Here’s the interesting part:

When we combined both tDCS and levodopa responses in a linear model, they jointly explained a significant amount of the variance in DBS outcomes.

So we asked, in our cohort:
(i) Does the levodopa response predict DBS outcome?
(ii) Does it correlate with the tDCS response?

The trends were there - but in this small sample (N = 10), neither reached significance on its own.

What did we find?

Levodopa improved motor function more when it followed multifocal tDCS than when it followed sham stimulation.

It seems that tDCS may prime the brain’s response to medication.

Luckily, in our cohort, most patients performed levodopa challenges on both days -
once after real tDCS, and once after sham stimulation.
So we took a closer look.

That brings us to the next part:
There’s long been debate about whether the levodopa challenge can reliably predict DBS outcomes.

Some studies say yes - others find the association limited.

This led us to a key question:
Do levodopa, tDCS, and DBS all act on the same brain network?

Some of those patients later underwent DBS - and interestingly, their DBS outcomes correlated with how much they had improved after tDCS.

A hint that both may tap into the same underlying network?
We think yes.

In 2024, we published a randomized, sham-controlled, double-blind study showing that multifocal tDCS, when targeted to a specific Parkinson's Disease response network, led to measurable motor improvements.

Publication in Movement Disorders:
doi.org/10.1002/mds....