Schematic overview of the specificity of brain perturbation techniques.

Brain perturbation techniques vary in the precision of the spatial and temporal effects that can be elicited, on logarithmic (log10) scales. This includes transcranial Focused Ultrasound Stimulation (TUS), in green. Some approaches with cellular specificity are shown that are currently primarily in use with nonhuman animals as models (optogenetics, infrared neuromodulation, chemogenetics and genetic manipulation). Figure modified with permission from P.C. Klink, from (Klink et al., 2021).

Low-intensity Transcranial Ultrasound Stimulation for Neuromodulation in Humans.

(A, B, C) Example focal TUS targeting of a human motor cortex using k-plan software (BrainBox, Inc.). (D) TUS simulation software uses an input set of parameters (e.g., pulse duration, sonication duration, pulse repetition frequency, transducer fundamental frequency, intensity in water (ISSPA)) to simulate and calculate the estimated TUS intensity in the target brain region using the participant’s MRI and CT scans if available, or template human brain and CT scans. Simulation software will also generate the complete set of minimal parameters for reporting.

Meta-analysis selection and inclusion criteria using the PRISMA recommended approach.

Selection and inclusion criteria for the meta-analysis, with resulting sample sizes for the meta-analysis.

Human ‘online’ TUS studies categorized by probable enhancement or suppression.

Summarized are the TUS parameters reported in the human studies (inTUS_DATABASE_1-2025.csv) and their reported neurobiological effects categorized by likely excitatory or inhibitory effects. For independently confirmed effects, we cite the independent assessment source. Abbreviations of Brain Targets: Somatosensory Cortex (S1,S2); Primary Visual Cortex (V1); Primary Motor Cortex (M1); Ventro-Lateral Prefrontal Cortex (VLPFC); Ventral Capsule/Ventral Striatum (VC/VS); dorsal-Anterior Cingulate Cortex (dACC); Orbito Frontal Cortex (OFC); Thalamus-Ventral Posterolateral Nucleus (VPL); human Middle Temporal Area (hMT+); Nucleus Accumbens (NAcc); Medial Prefrontal Cortex (mPFCF); Anterior Insula (AI); Posterior Insula (PI); Entorhinal Cortex (ErC); Abbreviations of Measurements: Heartbeat Evoked Potential (HEP); Contact Heat Evoked Potential (CHEP); Heart Rate Variability (HRV); Transcranial Magnetic Stimulation (TMS); Short-Interval Intracortical Inhibition (SICI); Intracortical Facilitation (ICF); rs-fMRI (Resting State functional Magnetic Resonance Imaging); 3 Tesla fMRI (3T); Blood Oxygenation Level Dependent (BOLD); Electro Encephalography (EEG).

Online and offline TUS study parameters and outcomes.

(A) Online effects, based on the studies in Table 1, are shown segregated by probable enhancement versus suppression. Light blue line shows the NICE model boundary between suppression and enhancement. Blue circles are human studies reporting probable enhancement, red circles probable suppression. Index numbers correspond to the studies numbered in Table 1. (B) Same format and analysis approach showing the “offline” effects studies in Table 2 with stars. Numbers correspond to the studies numbered in Table 2. (C) Combined figure with online (Table 1) and offline (Table 2) studies. Same symbol and color use as in A-B. (D) Additional hypothesized parameters, such as pulse repetition frequency (PRF), can be plotted from the inTUS_DATABASE in multi-dimensional spaces as shown and with the inTUS resource app.

Human ‘offline’ TUS studies categorized by probable enhancement or suppression.

Summarized are the TUS parameters used in human studies (inTUS_DATABASE_1-2025.csv) aiming to induce longer lasting ‘offline’ effects. Format as in Table 1. Abbreviations of Brain Targets: Dorso-Lateral Prefrontal Cortex (DLPFC); Primary Motor Cortex (M1); dorsal-Anterior Cingulate Cortex (dACC); Posterior Cingulate Cortex (PCC); Somatosensory Cortex (S1); Ventral Posterolateral Nucleus (VPL); Nucleus Accumbens (NAcc); Inferior-Prefrontal Gyrus (IPFC); Inferior Frontal Gyrus (IFG); right-Inferior Frontal Gyrus (rIFG); right-Medial Frontal Gyrus (rMFG). Abbreviations of Measurements: Transcranial Magnetic Stimulation (TMS); Short-Interval Interhemispheric Inhibition (SIHI); Intracortical Facilitation (ICF); Short-Interval Intracortical Inhibition (SICI); Short-Interval Intracortical Facilitation (SICF); Long-Term Potentiation (LTP); repetitive-Transcranial Ultrasound (rTUS); theta-burst TUS (tbTUS); resting-state functional Magnetic Resonance Imaging (rs-fMRI); Blood Oxygenating Level Dependent (BOLD); Electroencephalography (EEG); Motor Evoked Potential (MEP); Contact Heat Evoked Potential (CHEP); Reaction Time (RT); Heart Rate Variability (HRV); Default Mode Network (DMN).

Box plots of meta-analysis results for the TUS parameters of interest.

Shown are boxplots for TUS parameters of interest segregated by probable enhancement or suppression (data from Tables 1 and 2). Plots show TUS parameters: Pulse Repetition Frequency (PRF), Duty Cycle (DC), Sonication Duration (SD) and ISPPA Brain. These are shown separately for Online effects (A) and Offline effects (B).

Random Forest Walk multivariate pattern analysis results.

Results of the Random Forest analysis with the 4 parameters of interest in the second key database inTUS_DATABASE_1-2025 compared to a shuffled parameter dataset (left). Left panel shows the average accuracy (mean = 0.55; SD = 0.14) is not far from the shuffled dataset (mean = 0.48; SD = 0.15), suggesting that the dataset size may not yet be robust for multivariate analyses. The right panel shows the ordered decrease in accuracy when the variable is permuted (permutation importance: higher values carry more importance). We observe that the importance values are low (between –0.06 and 0.04; right panel), warranting caution in interpretation at this stage of the inTUS database