- TUS enables deep brain modulation but uncertainty in regional and cell type sensitivity limits optimal parameter selection and risks suboptimal treatments.
- Leveraging brain-wide gene expression from the Allen Brain Atlas and mechanistic gene sets can refine TUS parameter searches via dimensionality reduction and spatial clustering.
- Regional clusters of covarying gene expression predict similar TUS responsivity, explaining variable neuronal responses and emphasising region specific, molecularly informed parameter optimisation.
Imaging Neurosci (Camb). 2026 Jul 8;4:IMAG.a.1294. doi: 10.1162/IMAG.a.1294. eCollection 2026.
ABSTRACT
Non-invasive transcranial ultrasound stimulation (TUS) enables deep brain therapeutic exploration at unprecedented scale. However, its optimal use is limited by the uncertainty surrounding ultrasound sensitivity across brain regions and cell types. This uncertainty often forces selection of sub-optimal parameter-target treatment paradigms guided solely by precedent, rather than an unbiased search of the larger combinatorial space. In principle, human brain-wide gene expression data could allow for a refinement of the search space based on known mechanisms and associated gene expression. In this perspective, we discuss the practicality of genetically informed TUS parameter search in humans using the Allen Brain Atlas and incorporating a broad set of genes related to the hypothesized mechanisms of TUS neuromodulation. We define principal component expression patterns across the brain, enabling dimensionality reduction and spatial clustering of ultrasound-relevant gene expression data. We identify regional clusters of covarying gene expression profiles across the brain topology that are likely to have similar responsivity to TUS. These findings may explain previous perplexities around highly variant neuronal response across brain areas and highlight the need to optimize stimulation parameters in the context of brain region and its molecular profile.
PMID:42428552 | PMC:PMC13347599 | DOI:10.1162/IMAG.a.1294
Share Evidence Blueprint

Search Google Scholar
Save as PDF

