Neuromodulation with Ultrasound: Hypotheses on the Directionality of Effects and Community Resource

Curation statements for this article:
  • Curated by eLife

    eLife logo

    eLife Assessment

    This paper is an important overview of the currently published literature on low-intensity focussed ultrasound stimulation (TUS) in humans, providing a meta-analysis of this literature that explores which stimulation parameters might predict the directionality of the physiological stimulation effects. The overall synthesis, except for the section on TPS and AD, is convincing though could be streamlined at places. The database proposed by the paper has the potential to become a key community resource if carefully curated and developed.

This article has been Reviewed by the following groups

Read the full article See related articles

Abstract

Abstract

Low-intensity Transcranial Ultrasound Stimulation is a promising non-invasive technique for brain stimulation and focal neuromodulation. Research with humans and animal models has raised the possibility that TUS can be biased towards enhancing or suppressing neural function. Here, we first collate a set of hypotheses on the directionality of TUS effects and conduct an initial meta-analysis on the available healthy human participant TUS studies reporting stimulation parameters and outcomes (n = 47 studies, 52 experiments). In these initial exploratory analyses, we find that parameters such as the intensity and continuity of stimulation (duty cycle) with univariate tests show only statistical trends towards likely enhancement or suppressed of function with TUS. Multivariate machine learning analyses are currently limited by the small sample size. Given that human TUS sample sizes will continue to increase, predictability on the directionality of TUS effects could improve if this database can continue to grow as TUS studies more systematically explore the TUS stimulation parameter space and report outcomes. Therefore, we establish an inTUS database and resource for the systematic reporting of TUS parameters and outcomes to assist in greater precision in TUS use for brain stimulation and neuromodulation. The paper concludes with a selective review of human clinical TUS studies illustrating how hypotheses on the directionality of TUS effects could be developed for empirical testing in the intended clinical application, not limited to the examples provided.

Article activity feed

  1. eLife Assessment

    This paper is an important overview of the currently published literature on low-intensity focussed ultrasound stimulation (TUS) in humans, providing a meta-analysis of this literature that explores which stimulation parameters might predict the directionality of the physiological stimulation effects. The overall synthesis, except for the section on TPS and AD, is convincing though could be streamlined at places. The database proposed by the paper has the potential to become a key community resource if carefully curated and developed.

  2. Reviewer #1 (Public review):

    Summary:

    This paper is a relevant overview of the currently published literature on low-intensity focused ultrasound stimulation (TUS) in humans, with a meta-analysis of this literature that explores which stimulation parameters might predict the directionality of the physiological stimulation effects.

    The pool of papers to draw from is small, which is not surprising given the nascent technology. It seems nevertheless relevant to summarize the current field in the way done here, not least to mitigate and prevent some of the mistakes that other non-invasive brain stimulation techniques have suffered from, most notably the theory- and data-free permutation of the parameter space.

    The meta-analysis concludes that there are, at best, weak trends toward specific parameters predicting the direction of the stimulation effects. The data have been incorporated into an open database that will ideally continue to be populated by the community and thereby become a helpful resource as the field moves forward.

    Strengths:

    The current state of human TUS is concisely and well summarized. The methods of the meta-analysis are appropriate. The database is a valuable resource.

    Suggestions:
    - The paper remains lengthy and somewhat unfocused, to the detriment of readability. One can understand that the authors wish to include as much information as possible, but this reviewer is sceptical that this will aid the use of the databank, or help broaden the readership. For one, there is a good chunk of repetition throughout. The intro is also somewhat oscillating between TMS, tDCS and TUS. While the former two help contextualizing the issue, it doesn't seem necessary. In the section on clinical applications of TUs and possible outcomes of TUS, there's an imbalance of the content across examples. That's in part because of the difference in knowledge base but some sections could probably be shortened, eg stroke. In any case, the authors may want to consider whether it is worth making some additional effort in pruning the paper

    - The terms or concept of enhancement and suppression warrant a clearer definition and usage. In most cases, the authors refer to E/S of neural activity. Perhaps using terms such as "neural enhancement" etc helps distinguish these from eg behavioural or clinical effects. Crucially, how one maps onto the other is not clear. But in any case, a clear statement that the changes outlined on lines 277ff do not

    - Re tb-TUS (lines 382ff), it is worth acknowledging here that independent replication is very limited (eg Bao et al 2024; Fong et al bioRxiv 2024) and seems to indicate rather different effects

    - The comparison with TPS is troublesome. For one, that original study was incredibly poorly controlled and designed. Cherry-picking individual (badly conducted) proof-of-principle studies doesn't seem a great way to go about as one can find a match for any desired use or outcome.

    Moreover, other than the concept of "pulsed" stimulation, it is not clear why that original study would motivate the use of TUS in the way the authors propose; both types of stimulation act in very different ways (if TPS "acts" at all). But surely the cited TPS study does not "demonstrate the capability for TUS for pre-operative cognitive mapping". As an aside, why the authors feel the need to state the "potential for TPS... to enhance cognitive function" is unclear, but it is certainly a non-sequitur. This review feels quite strongly that simplistic analogies such as the one here are unnecessary and misleading, and don't reflect the thoughtful discussion of the rest of the paper. In the other clinical examples, the authors build their suggestions on other TUS studies, which seems more sensible.

  3. Author response:

    The following is the authors’ response to the original reviews

    Reviewing Editor Comments:

    Focus and Scope:

    The paper attempts to address too many topics simultaneously, resulting in a lack of focus and insufficient depth in the treatment of individual components.

    We have moved this selective clinical review section that was previously Part I in the paper now to Part II, given the importance of leading off with the meta-analysis and resource before doing a selective review, which are now Part I. In the lead in to Part II, we now indicate that the review is not intended to be comprehensive, because there are other recent comprehensive reviews, which we cite. This part of the paper merely aims to generate hypotheses on the directionality of effects ripe for testing on how TUS could be used to excite or suppress function, illustrated with specific clinical examples. The importance of this section, even though not comprehensive, is that it should provide the reader with examples on how the directionality of TUS could be used specifically in a range of clinical applications. The reader will find that the same hypotheses do not apply to different clinical disorder. Therefore, patient specific hypotheses need to be motivated and then subsequently tested with empirical application of TUS, which Part II provides.

    Part II. Selective TUS clinical applications review and TUS directionality hypotheses starts at line 458. Part I, the meta-analysis and resource section starts at line 199, after the Introduction on TUS and the importance on understanding how the directionality of TUS effects could be better understood.

    Strengthening the Meta-Analysis:

    The meta-analysis is the strongest aspect of the paper and should be expanded to include the relevant statistics. However, it currently omits several key concepts, studies, and discussion points, particularly related to replication and the dominance of results from specific groups. These omissions should be addressed even with a focus on meta-analysis.

    We thank the reviewer for their enthusiasm about the meta-analysis, which we have now promoted to Part I in the revised paper. We have substantially updated the latest database (inTUS_DATABASE_1-2025.csv) and ensured that the R markdown script can re-generate all of the results and statistical values. We have inserted additional statistical values in the main manuscript, as requested. The inTUS Resource is located here (https://osf.io/arqp8/ under Cafferatti_et_al_inTUS_Resource), and we have aimed to make it as user friendly to use and contribute to as possible. For instance, the reader can find them all in the HTML link summarizing the R markdown output with all statistical values here: https://rpubs.com/BenSlaterNeuro/1268823, a part of the inTUS resource.

    Since the last submission, there has been a tremendous increase in the number of TUS studies in healthy participants. We have curated and included all of the relevant studies we could find in the 1-2025 database, as the next large expansion of the database (now including 52 experiments in healthy participants). We then reran and report the results of the statistical tests via the R markdown script (starting at line 336). Finally, the online database (inTUS_DATABASE_1-2025.csv) has additional columns, suggested by the reviewers, including one to identify the same groups that conducted the TUS study, based on a social network analysis. The manuscript figures (Table 1 and Table 2) did not have the space to expand the data tables, but these additional columns are available in the database online. Finally, we have ensured that the resource is as easy to use as possible (line 862 has the Introduction to the inTUS Resource – which is also the online READ ME file), and we have been in contact with the iTRUSST consortium leads who are interested in discussing hosting the resource and helping it to become self-sustaining.

    Conceptual Development:

    The more conceptual part of the paper is underdeveloped. It lacks sufficient supporting data, a well-articulated argument, and a clear derivation or development of a concrete model.

    To ensure that the conceptual sections are well developed, we have revised the introduction, including the background on TUS and bases for the interest in the directionality of effects. We have also revised the TUS mechanisms background as suggested by the reviewers. For Part I, the meta-analysis basis and hypotheses we have ensured the rationale is clearer. The hypotheses are based on several lines of research in the animal model and human literature as cited (starting with line 211). For Part II, the selective clinical review, we have revised this section as well to have each section on lowintensity TUS and end in a hypothesis on the directionality of TUS effects. Starting at line 199 we have clarified the scope of the review and ensured that all the relevant experiments in healthy participants (n = 52 experiments) have now been included in the next key update of the resource and meta-analysis in this key paper update.

    Database Curation:

    The authors should provide more detailed information about how the database will be curated and made accessible. They may consider collaborating with ITRUSST.

    We have expanded the information on the Resource documents (starting at line 862) to make the resource as user friendly as possible. At the beginning of the resource development stage we had contacted but not heard from the ITRUSST consortium. Encouraged by this comment we again reached out and are now in contact with the ITRUSST consortium leads who are interested in discussing sustaining the resource. It would be wonderful to have the resource linked to other ITTRUST tools, since it was inspired by the organization. Practically what this means is that the resource rather than being hosted on Open Science Framework, would potentially be hosted on the ITRUSST web site (https://itrusst.com/). These discussions are in progress, but the next key update to the database (1-2025) is already available and reported in this key update to our original paper.

    Reviewer #1: (Public Review)

    Summary:

    This paper is a relevant overview of the currently published literature on lowintensity focussed ultrasound stimulation (TUS) in humans, with a meta-analysis of this literature that explores which stimulation parameters might predict the directionality of the physiological stimulation effects.

    The pool of papers to draw from is small, which is not surprising given the nascent technology. It seems nevertheless relevant to summarize the current field in the way done here, not least to mitigate and prevent some of the mistakes that other non-invasive brain stimulation techniques have suffered from, most notably the theory- and data-free permutation of the parameter space.

    The meta-analysis concludes that there are, at best, weak trends toward specific parameters predicting the direction of the stimulation effects. The data have been incorporated into an open database, that will ideally continue to be populated by the community and thereby become a helpful resource as the field moves forward.

    Strengths:

    The current state of human TUS is concisely and well summarized. The methods of the meta-analysis are appropriate. The database is a valuable resource.

    Weaknesses:

    These are not so much weaknesses but rather comments and suggestions that the authors may want to consider.

    We thank the reviewer for their support of the resource and meta-analysis. We have implemented the suggestions next as follows.

    I may have missed this, but how will the database be curated going forward? The resource will only be as useful as the quality of data entry, which, given the complexity of TUS can easily be done incorrectly.

    We have added a paragraph on how authors could use the Qualtrics form to submit their data and the curation process involved (from line 891). Currently, this process cannot be automated because we continue to find that reported papers do not report the TUS parameters that ITRUSST has encouraged the community to report (Martin et al., 2024). We can dedicate for a TUS expert to ensure that every 6 or 12 months the data base is curated and expanded. The current version is the latest 1-2025 update to the data base. Longer term we are in discussion with ITRUSST on whether the resource could become self sustaining when TUS papers regularly reporting all the relevant parameters such that the database expansion becomes trivial, and then the Resource R markdown script and other tools can be used to re-evaluate the statistical tests and the user can conduct secondary hypothesis testing on the data.

    It would be helpful to report the full statistics and effect sizes for all analyses. At times, only p-values are given. The meta-analysis only provides weak evidence (judged by the p-values) for two parameters having a predictive effect on the direction of neuromodulation. This reviewer thinks a stronger statement is warranted that there is currently no good evidence for duty cycle or sonication direction predicting outcome (though I caveat this given the full stats aren't reported). The concern here is that some readers may gallop away with the impression that the evidence is compelling because the p-value is on the correct side of 0.05.

    We have ensured that the R script can generate the full statistics from the tests and the effect sizes for all the analyses, and now also report more of the key statistical values in the revised paper (starting at line 336). As suggested, we have also ensured that the interpretation is sufficiently nuanced given the small sample sizes and the p-values below 0.1 but above 0.05 are interpreted as a statistical trend.

    This reviewer thinks the issue of (independent) replication should be more forcefully discussed and highlighted. The overall motivation for the present paper is clearly and thoughtfully articulated, but perhaps the authors agree that the role that replication has to play in a nascent field such as TUS is worth considering.

    We completely agree and have added additional columns to the online database to identify unique groups, using a social network analysis, and independent replications. These expanded tables did not fit in the manuscript versions of Tables 1 and 2 but are fully available in the Resource data tables ready for further analysis by interested resource users.

    A related point is that many of the results come from the same groups (the so-called theta-TUS protocol being a clear example). The analysis could factor this in, but it may be helpful to either signpost independent replications, which studies come from the same groups, or both.

    In the expanded database tables (inTUS_DATABASE_1-2025.csv: https://osf.io/arqp8/ under Cafferatti_et_al_inTUS_Resource) we have added a column to identify independent replication.

    The recent study by Bao et al 2024 J Phys might be worth including, not least because it fails to replicate the results on theta TUS that had been limited to the same group so far (by reporting, in essence, the opposite result).

    Thank you. We have added this study and over a dozen recent TUS studies in healthy participants to the database and redone the analyses.

    The summary of TUS effects is useful and concise. Two aspects may warrant highlighting, if anything to safeguard against overly simplistic heuristics for the application of TUS from less experienced users. First, could the effects of sonication (enhancing vs suppressing) depend on the targeted structure? Across the cortex, this may be similar, but for subcortical structures such as the basal ganglia, thalamus, etc, the idiosyncratic anatomy, connectivity, and composition of neurons may well lead to different net outcomes. Do the models mentioned in this paper account for that or allow for exploring this? And is it worth highlighting that simple heuristics that assume the effects of a given TUS protocol are uniform across the entire brain risk oversimplification or could be plain wrong? Second, and related, there seems to be the implicit assumption (not necessarily made by the authors) that the effects of a given protocol in a healthy population transfer like for like to a patient population (if TUS protocol X is enhancing in healthy subjects, I can use it for enhancement in patient group Y). This reviewer does not know to which degree this is valid or not, but it seems simplistic or risky. Many neurological and psychiatric disorders alter neurotransmission, and/or lead to morphological and structural changes that would seem capable of influencing the impact of TUS. If the authors agree, this issue might be worth highlighting.

    We agree that given the divergence in circuits and cellular constituents between cortical and subcortical areas, it is important to distinguish studies that have focused on cortical or subcortical brain areas. The online data tables identify the target region. The analyses can be used to focus on the cortical or subcortical sites for analysis, although for the current version of the database there are too few subcortical sites with which to conduct analyses on subcortical sites. On the second point, that pathology may have affected the results, we completely agree and have clarified that the current database only includes healthy participant experiments for this reason. We are considering future updates to the resource may include clinical patient results (Line 247).

    Reviewer #1 (Recommendations for the authors):

    Minor edits (I wouldn't call them "corrections").

    We sincerely appreciate the constructive comments and have aimed to address them all as suggested.

    Perhaps the most relevant edit pertains to the statistics.

    We now report the more complete statistical results (line 336) and the R markdown script can re-generate all the statistical values for the tests.

    The issue of replication also seems relevant and ought to be raised. This reviewer does not want to prescribe what to do or impose the view the authors ought to adopt.

    In the online version of the data tables for the latest dataset, we have added a column in the data table as suggested that identifies independent groups and replications.

    The other points are left to the authors' discretion.

    We have aimed to address all of the reviewer’s points. Thank you for the constructive input which has helped to improve the expanded database and resource.

    Reviewer #2: (Public Review)

    Summary:

    This paper describes a number of aspects of transcranial ultrasound stimulation (TUS) including a generic review of what TUS might be used for; a meta-analysis of human studies to identify ultrasound parameters that affect directionality; a comparison between one postulated mechanistic model and results in humans; and a description of a database for collecting information on studies.

    Strengths:

    The main strength was a meta-analysis of human studies to identify which ultrasonic parameters might result in enhancement or suppression of modulation effects. The meta-analysis suggests that none of the US parameters correlate significantly with effects. This is a useful result for researchers in the field in trying to determine how the parameter space should be further investigated to identify whether it is possible to indeed enhance or suppress brain activity with ultrasound.

    The database is a good idea in principle but would be best done in collaboration with ITRUSST, an international consortium, and perhaps should be its own paper.

    Weaknesses:

    The paper tries to cover too many topics and some of the technical descriptions are a bit loose. The review section does not add to the current literature. The comparison with a mechanistic model is limited to comparing data with a single model at a time when there is no general agreement in the field as to how ultrasound might produce a neuromodulation effect. The comparison is therefore of limited value.

    We appreciate the reviewer’s assessment and interest in the meta-analysis and database to guide the development of TUS for more systematic control of the directionality of neuromodulation. With this next key expansion of the database (inTUS_DATABASE_1-2025.csv) we have added over a dozen new studies that have been published since our original submission (n = 52 experiments). We have also moved the ‘review’ part of the paper below the meta-analysis and resource description. We have clarified that the clinical review section (now Part II in the revised manuscript) is not intended as a comprehensive review but as a selective review showing how hypotheses on the directionality of TUS effects need to be carefully developed for specific patient groups that require different effects to be induced at specific brain areas. Finally, we have gotten in contact with the ITRUSST consortium leads, as suggested, and are in discussion on whether the inTUS resource could be hosted by ITRUSST. Since these discussions are ongoing practically what this might mean is moving the resource from the Open Science Framework to ITRUSST webpages, which would be a trivial update of the link to the resource in OSF.

    We also sincerely appreciate the time and care the reviewer has given to provide us with the below guidance, all of which we have aimed to take on board in the revised paper.

    Reviewer #2 (Recommendations for the authors):

    Line 24/25 - I suggest avoiding using the term "deep brain stimulation" in reference to TUS as the term is normally used to describe electrically implanted electrodes.

    We have removed the term “deep” brain stimulation in reference to TUS to avoid confusion with electrical DBS for patient treatment [Line 24].

    Line 25 - I don't think "computational modelling" has changed how TUS can be done. There is still much to be understood about mechanisms. I think the modelling aspects of the paper should be toned down. Indeed the NICE data that is presented later appears to have a weak, if any, correlation to the outcomes.

    We have revised the manuscript text throughout to ensure that the computational modeling contributions are not overstated, as noted, given the lack of strong correlation to the NICE model outcomes by the meta-analysis including in the latest results with the more extensive database (n = 52).

    Line 32 - "exponentially increasing" is a well-defined technical term and the increase in studies should be quantified to ensure it is indeed exponential. I agree that TUS studies in humans are increasing but a quick tally of the data by year in the meta-analysis reported here doesn't suggest that it follows an "exponential" growth.

    We have changed “exponential” to “to increase”. [Line 32]

    Line 50 - I would suggest using the term sub-MHz rather than 100-1,000 kHz as it is challenging to deliver ultrasound at 1 MHz through the skull. The highest frequency in the meta-analysis is 850 kHz; but the majority are in the 200-500 kHz range.

    We have made this correction to sub-MHz. [Line 54]

    Line 58/59 - Is the FDA publication on diagnostic imaging relevant for saying that 50 W/cm2 is a lowintensity TUS? I think it's perhaps reasonable to say that intensities below diagnostic thresholds are "low intensities" but that is not clear in the text. I would refer to ITRUSST on what is appropriate for defining what is low, medium, or high.

    We have cut the reference to the FDA here since it is, as noted, not as relevant as pointing to the ITRUSST definition.

    Line 65/66 - I agree that ultrasound for neuromodulation is gaining traction and there is an increase in activity, but it also has a long history with the work of the Fry brothers published in the 1950s; and extensive work of Gavrilov in humans starting in the 1970s.

    We have added citations to the Fry brothers and Gavrilov to the text in this section. [Line 69/70]

    Line 75 - I think the intermembrane cavitation mechanism is unlikely to be due to "microbubbles" in a lipid membrane. The predicted displacements are on the order of nanometres, so they are unlikely to generate microbubbles. The work on comparing with NICE is limited. Note there are a number of experimental papers that have reported an absence of intra-membrane cavitation, including the Yoo et al 2022 which is referenced later in the paragraph. Also, there are other models, such as Liao et al 2021 (https://www.nature.com/articles/s41598020-78553-2).

    As suggested, we have removed this phrase on microbubble formation as a likely mechanism. We have also added the Liao paper to this paragraph as it is relevant.

    Line 83 - "At the lower intensities..." it is not clear whether this means all TUS intensities or the lower end of intensities used in TUS.

    We now use the following wording here: “low intensities”. [Line 86]

    Line 85/86 - "more continuous stimulation" the modulation paradigms haven't been described yet and so pulse vs continuous hasn't been made clear to the reader. Also "more continuous" is very loose terminology. Something is either continuous or it isn't.

    We agree and have removed “more” to be clear that the stimulation is continuous. [Line 88]

    Line 87/88 - "TUS does not .. cavitation ..when ..ISPTA...<14 W/cm2". You can't use ISPTA to determine cavitation. It is the peak negative pressure which is the key driver for cavitation and the MI which is the generally accepted (although grudgingly by some) metric for assessing cavitation risk. You can link the negative pressure to ISPPA but not really to ISPTA. In histotripsy for example the ISPTA is low due to the low duty cycles to avoid heating but the cavitation is a huge effect. Technical terminology is loose.

    We have corrected this to “TUS does not appear to cause significant heating or cavitation of brain tissue when the intensity remains low, based on Mechanical and Thermal Index values and recommendations of use”. [Line 90/91]

    Line 89 - What is meant by "low intensity TUS"? I think all TUS used in the literature counts as low intensity - in that it is below the level allowed for diagnostic imaging.

    We have ensured that the text is focused on TUS being low-intensity and only in the introduction do we distinguish low intensity TUS from moderate and high intensity TUS, such as used for thermal ablation [Lines 62-66].

    Line 88/89 - Most temperature rises in brain tissue in TUS are well below 1 C - will this really change membrane capacitance significantly? If so it would have been good to consider a model for it.

    We have revised this statement as “thermal effects could at least minimally alter cell membrane capacitance…”. [Line 93]

    Line 111 - The text refers to "recent studies" but then the next two references are from 1990 and 2005 which I would argue don't count as "recent".

    We have corrected this wording to “previous studies”. [Line 114]

    Lines 122/129 - This paragraph on TMS pulsing should be linked to the TUS paragraph on pulsing (lines 109/116). The intervening paragraph on anaesthesia is relevant but breaks the flow.

    We have merged the paragraph on anesthesia to the prior one on TUS so that the TMS paragraph is linked more closely to it [starting on line 112].

    Line 130/131 - It is not clear to me that current studies are being guided by computational models. I think there is still no generally accepted theory for mechanisms. If the authors want to do a mechanisms paper then they should compare a few.

    We have revised this as suggested to not overstate the contribution of the limited computational modeling studies throughout the manuscript.

    Line 132 on - There are a number of studies that suggest that NICE is likely not the mechanism by which TUS produces neuromodulation.

    We have revised this sentence as follows: “Although it remains questionable whether intramembrane cavitation is a key mechanism for TUS, the NICE model simulations explored a broad set of TUS parameters, including TUS intensity and the continuity of stimulation (duty cycle) on modelled neuronal responses.” [Lines 139/142]

    Lines 137-140 - Terms are defined after their use. Things like ISPPTA, PRF, TI, and MI have been discussed already and so the terms should have been defined earlier. The authors should think carefully about how the material is presented to make it more logical for the reader.

    We have ensured that the definitions precede the use of abbreviations and have added abbreviations to the tables.

    Part I Line 180-437 - The review of potential applications for TUS reads like an introductory chapter of a thesis. It is entirely proper for a thesis to have a chapter like this, but it is not really relevant for a peer-reviewed research article. There are also numerous applications, e.g. mapping areas associated with decisions, or treating patients with addiction, which are not included, so it is not exhaustive. I would suggest this part be removed.

    We have moved the ‘review’ part of the paper to Part II, given the metaanalysis and resource should be more prominent as Part I. In the review now Part II of the paper we also now make it clear that there are recent comprehensive reviews of the clinical literature ( line 465/467). Namely, the purpose of our selective review is to demonstrate how directionality of TUS effects need to be specific for the clinical application intended, given the great variability in clinical effects that might be desired, brain areas targeted and pathology being treated. We have also aimed to ensure that each section summary is scholarly and academically written to a high level. All the co-authors contributed to these sections so we have also edited to have some consistency across sections, with sections ending with directionality of TUS hypotheses that could be developed for empirical testing.

    Line 453 - It is stated that "ISPTA, which mathematically integrates ISSPA by the sonication DC" It sounds rather grand to mathematically integrate but you can't integrate with respect to DC, you can integrate with respect to time. If you integrate intensity with respect to time over pulse and over the sonication time then one finds that ISPTA = DC x ISPPA, multiplication is also an important mathematical function and should be given its due. Lastly, I think there is a typo and ISSPA should read ISPPA

    We have corrected the typo and the statement to “mathematically multiplies ISPPA by the continuity of sonication”. [Line 221/222]

    Line 454 - I don't think ISPTA is a good measure of "dose." In radiation physics dose is well defined in terms of absorbed energy. The equivalent has yet to be defined for TUS so I would avoid using dose. The ISPTA does relate to TI - although it depends not just on the spatial peak but also on the spatial distribution and the frequency-dependent absorption coefficient of the tissue. I would just avoid the use of "dose" until the field has a better idea of what is going on.

    We have cut this phrase on dose as suggested.

    Page 16 Box 1 - TI is defined as diagnostic ultrasound imaging it is based on. Also, I think TI is dimensionless; it is referenced to a 1-degree temperature rise and so it can be interpreted in terms of celsius or kelvin; but to be technically accurate it is dimensionless.

    We have made TI dimensionless in Box 1

    Page 17 Box 2 - Here you have no units for TI - which is correct but inconsistent with Box 1. But the legend suggests a 2 K temperature rise where as your Box allows for 6 K. The value of 6 is consistent with FDA but my understanding of the BMUS guidelines is the TI must be less than or equal to 0.7 for unlimited time or less than 3 if the duration is less than 1 minute. I accept that the table is labelled FDA limits, but the bold table caption is "Recommendations for TUS parameters" I think you should give the ITRUSST values rather than FDA.

    We have revised this Box legend to better distinguish the FDA and ITRUSST recommendation where they differ (e.g., the importance of ISPTA and the TI values). See revised legend for Box 2.

    Page 18 Box 3 - Not sure what this is trying to show? Also, what is "higher intensity" and "lower intensity"?

    Why not just give a range of values in each box?

    We agree that the higher and lower intensities likely to lead to enhancement or suppression are poorly defined and have noted this in the legend: “Note that the threshold for ISPPA qualifying as ‘higher’ or ‘lower’ intensity is currently poorly understood, or may non-linearly interact with other factors” [Line 751/754, Box 3].

    Line 444 - The hypotheses should be stated more clearly. Maybe I am just dense, but it is not obvious to me from box 3.

    We provide the basis for the hypotheses in the manuscript text on the paragraph [Lines 106-179].

    Line 481/482 - The intensity of a diagnostic ultrasound system is very well characterised. It just might be that the authors didn't report it. It is not clear what is meant by the "continuity." I guess it's to do with pulsing - which is also well defined but perhaps also not reported.

    We agree and have revised this as follows “For the meta-analysis, we only included studies that either reported a basic set of TUS stimulation parameters or those sufficient for estimating the required parameters or those sufficient for estimating the required parameters necessary for the meta-analysis” [Lines 256/258]

    Figure 2 - What is the purpose of this figure? Did you carry out simulations for all the studies? It doesn't seem to be relevant to the data here.

    This figure illustrates the TUS targeting approach and simulations, in this case conducted in k-plan. These were conducted to evaluate approximations to ISPPA in brain values from the studies that did not report these values [Lines 264/268]).

    Figure 4 - The data in these figures is nice (and therefore doesn't need to have a NICE curve) To me it clearly shows that the data in the literature does not obviously segment into enhancement vs suppression with DC. I suspect it is the same with PRF. I think it would have been better if C and D had PRF on the horizontal axis for on-line and off-line so that effect could be seen more clearly.

    We have kept the NICE curve only for a reference that some readers familiar with the NICE model might want to see overlaid in the figure, but have ensured that the text throughout makes clear that the NICE model predictions are not as statistically robust as initially anecdotally thought. PRF results are not significant but we do show a panel with the PRF measures on one axis (Fig. 4D). Figure 5 also shows box plot results with PRF as well as the other key TUS parameters. Moreover, in the inTUS resource we have provided an app for users to explore the data (https://benslaterneuro.shinyapps.io/Caffaratti_inTUS_Resource/).

    Figure 5 - The text on the axes is too small to read. Was the DC significant for both on-line and offline? What about ISPPA for off-line. At least by eye, it looks as different as DC. Figure 5C doesn't add anything.

    We have boosted the font for Figure 5 and have cut panel 5C since it was not adding much. We have also checked whether DC parameter was significant separately for on-line and off-line effects, but the sample sizes were too small for significance, and the statistical test was not significantly different for Online and Offline effects even in the 12025 database. Therefore they might look stronger for Offline effects in some of the plots in Figure 5, but are currently statistically indistinguishable [Lines 347/348].

    Table 1 - There is a typo in the 3rd column. FF should have units of kHz, not KHz. In addition, SD should have units of s as that is the SI symbol for seconds. I would swap columns 9 and 10 so that ISPPA in water and ISPPA in the brain are next to each other.

    We have corrected the typo in the 3rd column and ensured that units are kHz. SD in the tables has units of ‘s’ for seconds and have put ISPPA in water and in brain next to each other in the data tables.

    Line 767 - "M.K. was supported..." There are TWO MKs in the author list.

    We have changed this to M.Ka. for Marcus Kaiser.

  4. eLife Assessment

    This paper is an important overview of the currently published literature on low-intensity focussed ultrasound stimulation (TUS) in humans, with a meta-analysis of this literature that explores which stimulation parameters might predict the directionality of the physiological stimulation effects. Whilst currently incomplete, the database proposed by the paper has the potential to become a key community resource if carefully curated and developed.

  5. Reviewer #1 (Public review):

    Summary:

    This paper is a relevant overview of the currently published literature on low-intensity focussed ultrasound stimulation (TUS) in humans, with a meta-analysis of this literature that explores which stimulation parameters might predict the directionality of the physiological stimulation effects.

    The pool of papers to draw from is small, which is not surprising given the nascent technology. It seems nevertheless relevant to summarize the current field in the way done here, not least to mitigate and prevent some of the mistakes that other non-invasive brain stimulation techniques have suffered from, most notably the theory- and data-free permutation of the parameter space.
    The meta-analysis concludes that there are, at best, weak trends toward specific parameters predicting the direction of the stimulation effects. The data have been incorporated into an open database, that will ideally continue to be populated by the community and thereby become a helpful resource as the field moves forward.

    Strengths:

    The current state of human TUS is concisely and well summarized. The methods of the meta-analysis are appropriate. The database is a valuable resource.

    Weaknesses:

    These are not so much weaknesses but rather comments and suggestions that the authors may want to consider.

    (1) I may have missed this, but how will the database be curated going forward? The resource will only be as useful as the quality of data entry, which, given the complexity of TUS can easily be done incorrectly.

    (2) It would be helpful to report the full statistics and effect sizes for all analyses. At times, only p-values are given. The meta-analysis only provides weak evidence (judged by the p-values) for two parameters having a predictive effect on the direction of neuromodulation. This reviewer thinks a stronger statement is warranted that there is currently no good evidence for duty cycle or sonication direction predicting outcome (though I caveat this given the full stats aren't reported). The concern here is that some readers may gallop away with the impression that the evidence is compelling because the p-value is on the correct side of 0.05.

    (3) This reviewer thinks the issue of (independent) replication should be more forcefully discussed and highlighted. The overall motivation for the present paper is clearly and thoughtfully articulated, but perhaps the authors agree that the role that replication has to play in a nascent field such as TUS is worth considering.

    (4) A related point is that many of the results come from the same groups (the so-called theta-TUS protocol being a clear example). The analysis could factor this in, but it may be helpful to either signpost independent replications, which studies come from the same groups, or both.

    (5) The recent study by Bao et al 2024 J Phys might be worth including, not least because it fails to replicate the results on theta TUS that had been limited to the same group so far (by reporting, in essence, the opposite result).

    (6) The summary of TUS effects is useful and concise. Two aspects may warrant highlighting, if anything to safeguard against overly simplistic heuristics for the application of TUS from less experienced users. First, could the effects of sonication (enhancing vs suppressing) depend on the targeted structure? Across the cortex, this may be similar, but for subcortical structures such as the basal ganglia, thalamus, etc, the idiosyncratic anatomy, connectivity, and composition of neurons may well lead to different net outcomes. Do the models mentioned in this paper account for that or allow for exploring this? And is it worth highlighting that simple heuristics that assume the effects of a given TUS protocol are uniform across the entire brain risk oversimplification or could be plain wrong? Second, and related, there seems to be the implicit assumption (not necessarily made by the authors) that the effects of a given protocol in a healthy population transfer like for like to a patient population (if TUS protocol X is enhancing in healthy subjects, I can use it for enhancement in patient group Y). This reviewer does not know to which degree this is valid or not, but it seems simplistic or risky. Many neurological and psychiatric disorders alter neurotransmission, and/or lead to morphological and structural changes that would seem capable of influencing the impact of TUS. If the authors agree, this issue might be worth highlighting.

  6. Reviewer #2 (Public review):

    Summary:

    This paper describes a number of aspects of transcranial ultrasound stimulation (TUS) including a generic review of what TUS might be used for; a meta-analysis of human studies to identify ultrasound parameters that affect directionality; a comparison between one postulated mechanistic model and results in humans; and a description of a database for collecting information on studies.

    Strengths:

    The main strength was a meta-analysis of human studies to identify which ultrasonic parameters might result in enhancement or suppression of modulation effects. The meta-analysis suggests that none of the US parameters correlate significantly with effects. This is a useful result for researchers in the field in trying to determine how the parameter space should be further investigated to identify whether it is possible to indeed enhance or suppress brain activity with ultrasound.

    The database is a good idea in principle but would be best done in collaboration with ITRUSST, an international consortium, and perhaps should be its own paper.

    Weaknesses:

    The paper tries to cover too many topics and some of the technical descriptions are a bit loose. The review section does not add to the current literature. The comparison with a mechanistic model is limited to comparing data with a single model at a time when there is no general agreement in the field as to how ultrasound might produce a neuromodulation effect. The comparison is therefore of limited value.