double-blind trial comparing Beam-F3 and our personalized targets.

Thank you to all the co-authors @leonooi.bsky.social @shansiddiqi.bsky.social @foxmdphd.bsky.social @twktan.bsky.social who made this possible, and the patients who participated in the trial.

hospital historically achieved response rate of 21%, indicating a patient population profile less responsive to TMS than those recruited in typical clinical trials. But of course, this is open label, and placebo effects are likely stronger with a cobot, so we are running a ...

used to estimate personalized targets for other psychiatric conditions (e.g., anxiety target) given a target map, such as those produced by @shansiddiqi.bsky.social @foxmdphd.bsky.social

In an open label trial, TAO-TMS achieves a 70% response rate. For context, non-accelerated BeamF3 TMS at the...

approaches. Being closer to scalp allows the reduction of stimulation intensity, which can reduce patient headaches. Compared with the cone and cluster algorithms, our TAO-TMS targets enjoy better network specificity and engagement of subgenual ACC. Our approach can also be ...

Spatial Topography of Individual-Specific Cortical Networks Predicts Human Cognition, Personality, and Emotion Abstract. Resting-state functional magnetic resonance imaging (rs-fMRI) offers the opportunity to delineate individual-specific brain networks. A major que

builds upon the MS-HBM network mapping algorithm previously developed by Ruby (doi.org/10.1093/cerc...), and ensures the target is close to the scalp. We have previously demonstrated that MS-HBM can estimate high-quality brain networks with much shorter scan time than other

We develop a new TMS targeting algorithm and test it in an open label trial in a treatment-resistant depression population with high comorbidities. Preprints by @rubykong92.bsky.social Phern-Chern Tor
1. doi.org/10.1101/2025...
2. doi.org/10.64898/202...

Our new approach ...

with simulated data does not inform actual use case of fitting to real data. This simulation-reality gap is why simulation based inference does not perform well (Figure S4). 2/2

Great question! We didn't test parameter recovery because multiple parameter configs produce equally realistic dynamics, so even numerical integration might not recover ground truth parameters. More importantly, these models only approximate real human fMRI, so successful parameter recovery ... 1/2

revealing new insights into sex differences and network-specific patterns.

Thank you to the lead authors Tianchu Zeng and @tianfang.bsky.social and the many co-authors who made this possible. @shaoshiz.bsky.social @bart-larsen.bsky.social @ted-satterthwaite.bsky.social @avramholmes.bsky.social

Updated preprint: doi.org/10.1101/2025...

We have improved DELSSOME and showed that we can accelerate the estimation of two new biophysical models. By collating 12,005 individuals, we derive normative trajectories of cortical E/I ratio across the lifespan ...

Hi Brian, in Trevor’s preprint we do include pseudo respiratory motion filtering in the main analysis.

What are the downstream implications of censoring high motion volumes in fMRI? Two new preprints find that aggressive censoring leads to noisier FC, more attenuated and variable BWAS, and worse personalized TMS targets.

arxiv.org/html/2603.07...
www.biorxiv.org/content/10.6...

Thanks to all co-authors Damon Pham, Joanne Hwang, @nichols.bsky.social @sneuroble.bsky.social @ted-satterthwaite.bsky.social @twktan.bsky.social @rubykong92.bsky.social @shansiddiqi.bsky.social and especially @bttyeo.bsky.social for being an indefatigable partner in crime!

Thanks to all the co-authors @twktan.bsky.social @rubykong92.bsky.social @shansiddiqi.bsky.social @nichols.bsky.social @sneuroble.bsky.social @ted-satterthwaite.bsky.social who made this possible, especially @mandymejia.bsky.social who initiated this whole crusade.

/end

common dilemma is whether to re-scan a patient with high motion. These results suggest that within the upper bound of motion explored in our study (max frame-level FD after pseudo-motion filtering: 1.2mm or max FDrms: 1.1mm), a re-scan is not needed for personalized TMS.

whether the resulting FC or TMS target is close to the ground truth FC or TMS targets. In general, we find that lenient or even no censoring can lead to better outcomes.

These studies have important implications for personalized TMS, which is increasingly used in the clinic. A

Mandy's idea is to use precision-fMRI where each individual has large amount of data. We then use 1 or 1.5 hour of very clean data to define the ground truth FC or ground truth TMS targets.

In separate scans from the same individual, we can then vary the censoring level to see

In this plot, notice how the within-subject QC-FC is much weaker than between-subject QC-FC, suggesting that much of the QC-FC correlations we observe is motion traits, as opposed to artifacts.

How then can we determine the best censoring level?

A hint comes from the seminar work by Hesheng Liu, which shows that much of the motion artifacts might actually be motion traits, not noise. www.pnas.org/doi/abs/10.1...

By applying QC-FC to repeated scans, where motion level is different across the scans, we can tease out motion traits vs state,

Indeed, Mandy's preprint shows that FC error increases with more censoring.

How is that possible given that many studies have shown that motion is correlated with motion and censoring reduces such correlation, i.e., QC-FC plots by the great Jonathan Power?

Do you censor high motion frames in fMRI? In two preprints by @twktan.bsky.social @mandymejia.bsky.social, we find that we may be censoring too much!

doi.org/10.64898/202...
arxiv.org/html/2603.07...

Strict censoring leads to worse personalized TMS targets than no censoring, even with high motion!

Neuroanatomy reflects individual variability in impulsivity in youth - Molecular Psychiatry Molecular Psychiatry - Neuroanatomy reflects individual variability in impulsivity in youth

Our latest work looking at the neuroanatomical basis of impulsivity in youth is out now in Molecular Psychiatry!

FDA refuses to review Moderna’s application for mRNA flu vaccine, company says | CNN The US Food and Drug Administration has refused to accept an application from Moderna to review its first mRNA seasonal flu vaccine, the company said Tuesday, in another setback for the technology tha...

edition.cnn.com/2026/02/10/h...

Important study!

Thanks Nicole!

Preprint is now published! doi.org/10.1002/hbm....

Thanks to co-authors @chen-zhang.bsky.social @anlijuncn.bsky.social @csabaorban.bsky.social

Preprint is now published: doi.org/10.1002/hbm....

Special thanks to co-authors @twktan.bsky.social @kimnganngt.bsky.social @csabaorban.bsky.social

Excited to see this one in print! Emerging methods for brain-based predictions in psychiatry. Led by @diawang.bsky.social and @yiplab.bsky.social w/ @bttyeo.bsky.social

From big to small: Emerging methods for enhancing precision psychiatry through transfer learning Identifying reliable links between individual differences in neurobiological features and differences in symptom profiles or treatment outcomes is a primary goal of precision psychiatry. Within this c...

Excited to share this review in @biologicalpsych.bsky.social on using transfer learning to leverage large MRI datasets & enhance precision psychiatry!!

www.biologicalpsychiatryjournal.com/article/S000...

Fun collab with rockstars @diawang.bsky.social @bttyeo.bsky.social @avramholmes.bsky.social