Scientists’ perspectives on ethical issues in research with emerging portable neuroimaging technology: The need for guidance on ethical, legal, and societal implications (ELSI)

Abstract

Deployment of new, more portable, and less costly neuroimaging technologies such as portable magnetoencephalography, electroencephalography, positron emission tomography, functional near-infrared spectroscopy, high-density diffuse optical tomography, and magnetic resonance imaging is advancing rapidly. Given this trajectory toward increasing use of neuroimaging outside the hospital, we sought to identify ethical, legal, and societal implications (ELSI) of these new technologies by understanding the perspectives of those scientists and engineers developing and implementing portable neuroimaging technologies in the United States, Europe, and Asia. Based on a literature review, we identified and contacted 19 potential interviewees and then conducted 11 semi-structured interviews in English by Zoom. Analysis of the interviews revealed key themes and ELSI issues. Developers reported that without proper ELSI guidance, portable and accessible neuroimaging technology could be misused, fail to comply with applicable regulation and policy, and ultimately fall short in its mission to provide neuroimaging for the world. Our interviews suggested that ELSI guidance should address differences between imaging modalities because they vary in capability, limitations, and likelihood of generating incidental findings.

1 |. INTRODUCTION

Portable neuroimaging technologies are rapidly proliferating, ushering in a fundamental shift in how neuroimaging research will be performed by allowing for field-based research in real-world settings.¹ In recent years, portable low-field magnetic resonance imaging (pMRI) scans have been conducted for the first time in a moving ambulance, in a van outside a house, and at the bedside in an ICU.² Researchers are using portable electroencephalography (EEG) to monitor students in class and truck drivers on the road, deploying functional near-infrared spectroscopy (fNIRS) to acquire brain data while humans perform Tai Chi exercise and in the field to measure the effects of pesticide exposure on the brain, and utilizing high-density diffuse optical tomography (HD-DOT) in low-resource settings to study the effects of malnourishment on the brain.^3–7 In addition, portable high-field MRI, portable positron emission tomography (PET), and portable magnetoencephalography (MEG) are all under development.^8–10 Build-it-yourself low-cost MRI projects are also underway.^11,12

While these different imaging modalities vary in size, expense, portability, spatial and temporal resolution, and other technical details,¹³ they share the potential to facilitate neuroimaging research with historically marginalized populations, population neuroscience research with large-N samples in community settings, and innovative research designs outside a fixed scanning facility.¹⁴ Portable neuroimaging technology thus “has transformative potential across a range of neuroscientific and clinical applications.”⁹ Portable MRI—one type of portable neuroimaging technology—may reach even further; the recent surge of interest in direct-to-consumer (DTC) whole body MRI to screen for health issues¹⁵ suggests that commercial ventures such as wellness spas featuring neurofeedback or brain-health clubs providing annual scans for members may consider incorporating pMRI as well.¹⁶

While portable neuroimaging can be used in both clinical and research settings, our focus in this article is on research use. By facilitating collection of neuroimaging in new locations, and by lowering the barriers of entry to neuroimaging research, portable neuroimaging will allow novel research designs featuring data acquisition in the real world and new types of researchers—including local communities and citizen scientists.¹

The rapid emergence of pMRI and additional portable neuroimaging technologies give rise to urgent ethical, legal, and societal issues (ELSI). These include how to protect privacy in new settings, secure informed consent in underserved populations, manage incidental findings far from a medical center, exercise control over data acquired remotely, avoid bias when applying data analytic tools in new populations, and offer return of results.^1,17 The term ELSI emerged in the 1980s out of the US NIH-DOE Human Genome Project (HGP), to address “the challenge of developing ways to anticipate ethical, legal, and social implications before science provokes them as practical policy issues.”¹⁸ What started out as a single ELSI Working Group within the HGP has now blossomed into “a whole field of ‘ELSI research’ with its own academic infrastructure of institutional centers, training programs, consortia, conferences, and international networks.”^18,19 ELSI research in genetics and genomics spurred ELSI research in neuroscience.^20–22

This article grows out of a larger ELSI project supported by the National Institutes of Health (NIH) Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative (Highly Portable and Cloud-Enabled Neuroimaging Research: Confronting Ethics Challenges in Field Research with New Populations, NIH RF1MH123698). The project investigators, collaborating with an interdisciplinary Working Group, “identified 15 core pMRI research ELSI issues along with recommended solutions.”²³ While the Working Group’s recommendations are designed for pMRI, the substance of that guidance can be adapted for additional highly portable imaging modalities such as EEG, MEG, and fNIRS.

Although our research supported by the NIH BRAIN grant was focused specifically on pMRI research, to inform this process we initially sought guidance from those developing and deploying a range of portable neuroimaging technologies, including but not limited to pMRI, in order to place our focal ELSI issues in context. Multiple approaches to analyzing emerging technologies—including “anticipatory governance,” “responsible research and innovation,” and guidance from the Committee on Emerging Science, Technology, and Innovation in health and medicine (CESTI) of the National Academies of Science, Engineering, and Medicine (NASEM) recommend conducting empirical analysis to anticipate the ELSI issues before a technology is widely deployed.^24–30 One method used involves conducting interviews with key stakeholders to identify major issues and potential solutions.³¹ Adopting this approach, we conducted 11 semi-structured interviews with scientists and engineers who are developing and implementing a range of portable brain scanning technologies, including portable MRI, EEG, fNIRs, and PET.

We interviewed respondents from the United States, Europe, and Asia, though most interviewees were located in and based their research in the United States. As the technology being developed by our US interviewees may be deployed across the world, we probed for ethical and legal issues generally without focusing on a particular country’s regulations in the interviews. However, our discussion in this article of anticipated ELSI issues grounds our analysis in US law and regulation, given that most of our interviewees were US-based.

2 |. METHODS

Our stakeholder scientist interviews were designed to help inform anticipatory governance for pMRI by identifying scientists’ views of potential applications of portable neuroimaging technologies and of the key ELSI issues. Qualitative research using semi-structured interviews is an established method in empirical bioethics.³¹ In neuroethics, this method has been used to address a variety of research questions, including stakeholder perspectives on the use of neuroimaging in mental health care,³² the use of neuroimaging for attention disorders,³³ and a range of ethical issues in deep brain stimulation research.^34–36 Stakeholder interviews can “help identify ethical and policy issues for further analysis.”³⁴

In related articles, we have reported our use of complementary methods for assessing the views of both experts and the general public on ELSI issues associated with pMRI research. In one study, we utilized a nationally representative survey of the US public to assess the general public’s willingness to participate in pMRI research and their assessment of potential benefits and concerns.³⁷ In an additional study, we surveyed 114 expert stakeholders in neuroscience, ethics, and related fields to assess their views on ELSI challenges and solutions associated with pMRI research.³⁸

The survey methods we used in those articles allowed for quantitative analysis of larger N datasets, revealing group and sub-group trends. However, surveys do not permit the in-depth probing that interview methodology can provide. The interviews reported in this article informed the survey studies by helping to generate an initial roster of ELSI issues for inclusion on the survey.

Before conducting the semi-structured interviews reported in this article, we preliminarily identified key issues through conversations in December 2018 with several neuroimaging experts who were collaborating with us on the NIH BRAIN neuroethics supplement (3U01EB025153–02S2) that in part supported this research. A literature review, initially conducted between November 2018 through March 2020 and then updated periodically after, was used to identify key stakeholders to be considered for potential interviews.

From August 2019 through August 2021, we conducted 11 semi-structured 30-min Zoom interviews with scientists developing portable neuroimaging technologies in the United States (8), Europe (2), and Asia (1). We defined our relevant study population as those fluent in English who had published on the development or deployment of a field-based neuroimaging technology. In addition, snowball sampling was utilized by asking each interviewee at the end of the interview to recommend additional potential scientists to interview. Snowball sampling is a method in which the investigator asks initial participants to recommend others who may be interested in and eligible for the study, in this case asking for recommendations of others who are developing and utilizing portable neuroimaging.³⁹ Such a method has been recognized as particularly useful for accessing hard-to-reach populations,⁴⁰ as in this case because the universe of portable neuroimaging developers was quite small in 2019. Prospective participants were invited by email to undergo the interview. The email contained a consent and information sheet, and this information was circulated again in advance of the interview.

We contacted 19 potential interviewees, of which 11 agreed to participate. The other eight either did not reply or declined our invitation to the interview. Interviewees’ primary occupation type and professional training were coded by a co-author (F.S.) based on analysis of their publications and online biography and CV where available (Table 1). The interviewees were mainly academic researchers, with a balance of engineers and neuroscientists. Seven of the 11 interviewees were primarily involved in building the machines, 2 had scientific backgrounds but were now primarily bringing portable neuroimaging to market in industry, and 2 used portable neuroimaging in their research. The interviewees had expertise across multiple modalities of portable neuroimaging (Table 1). We note that a limitation of the study is that we interviewed only those who were actively engaged in developing and deploying portable neuroimaging. Given their relationship to portable neuroimaging, it can be expected that they may be more enthusiastic or confident about the potential of these technologies. Future studies could expand the interview pool to include experts who might be more skeptical about the promise of portable neuroimaging.

TABLE 1.

Professional backgrounds of experts interviewed

Interviewee backgrounds (N = 11)	N (%)
1. Primary occupation typea
Academics	7 (64%)
Physicians	1 (9%)
Private-sector employees	3 (27%)
2. Professional traininga
Engineers	4 (36%)
Neuroscientists	4 (36%)
Physicists	2 (18%)
Physicians	1 (9%)
3. Areas of expertiseb
Portable MRI	4 (36%)
Portable EEG	4 (36%)
Portable fNIRS	2 (18%)
Portable HD-DOT	1 (9%)
Portable MEG	1 (9%)

Abbreviations: EEG, electroencephalography; fNIRS, functional near-infrared spectroscopy; HD-DOT, high-density diffuse optical tomography; MEG, magnetoencephalography; MRI, magnetic resonance imaging.

^aPrimary occupation and professional training were ascertained by F.S. based on online CV, online biography, and publications.

^bTotal percentage for “areas of expertise” exceeds 100 because one respondent identified more than one area of expertise.

All 11 participants consented to having the interview recorded. De-identified transcripts were created and used for analysis and reporting of results, with the recordings destroyed. The University of Minnesota Institutional Review Board determined that this research met the criteria for exemption from IRB review (STUDY00006936). The 11 interviews were conducted by a co-author (F.S.).

Interviews were conducted using a semi-structured interview guide (see the supporting information). As the first study to probe these ELSI issues with portable neuroimaging developers, this research was exploratory, with a goal of identifying emerging themes that can be interrogated further in future investigations. The interviews explored (1) the general scientific and research landscape in which the participant’s particular technology fits; (2) the specifics of the participant’s particular technology; (3) envisioned use cases across research, clinical care, and commercial uses; and (4) the participant’s views of the potential ELSI implications of the technology. The first three segments of the interview provided our research team with a non-technical introduction to these new technologies and the participant’s view of how they are likely to be deployed. The rest of the interview invited the participant to identify ELSI issues and lastly to suggest other relevant stakeholders with whom we should engage. We report in this article primarily on the probable use cases and ELSI components of the interviews.

Interviews were recorded on Zoom and transcribed by a research assistant (RA). Interview transcripts were then de-identified by a senior member of the research team (F.S.). A different member of the research team (F.L.) plus two different RAs (F.D., E.T.) performed a qualitative analysis, coding themes in 5 of the 11 de-identified interview transcripts. The three coders then compared results as the basis for discussion to refine the codes and resolve disagreement by consensus. Each of the remaining six interviews were coded by one RA (three by F.D. and three by E.T.) with subsequent discussion and more refinement of the codes to ensure consistency and accuracy. Once all interview transcripts were coded, two members of the research team (F.S. and F.L.) and the two RAs met to extract themes. This yielded categories across three overlapping areas: state of the science, ethical issues, and the researcher-participant relationship. The sub-categories related to ethical issues included overpromising what the technologies can currently do and how to ensure privacy of acquired brain data. The sub-categories related to the researcher–participant relationship included how researchers should converse with participants when seeking informed consent and how to manage incidental findings. Organizing the interview data into these primary categories and sub-categories facilitated discussion by all authors and identification of the most relevant findings for purposes of understanding how ELSI issues are arising and being addressed in research utilizing portable neuroimaging technologies.

3 |. KEY FINDINGS

We identified four main findings (Table 2). First, our interviews found that the scientific stakeholders developing portable neuroimaging technologies have little experience with ELSI issues and none of the interviewees reported engaging in analysis of ELSI issues during the neurotechnology development phase. There is a clear need to spur deeper engagement with ELSI for technology developers. Second, our interviews suggest that many of these technologies are being designed not just as stand-alone devices but rather as part of what we call a “stacked” set of technologies that rely heavily on the additional technology of AI. Guidance on ELSI issues for portable neuroimaging thus should integrate guidance on ethical and trustworthy artificial intelligence (ETAI). Third, our interviews confirm that portable neuroimaging devices are being engineered for nonexpert users in remote locations, raising concerns around the capacity and requisite training that these teams will have to interpret and utilize the information generated. Fourth, although some general ethical principles will apply across all modalities of portable neuroimaging, ELSI guidance will need to address specific imaging modalities, because our interviews showed that the salience of specific ELSI issues varies by technology.

TABLE 2.

Key findings arising from stakeholder interviews

#	Findings	Recommended solutions
1	None of the 11 interview participants were aware of, or had applied, ELSI guidance addressing development or deployment of portable neuroimaging technologies.	Researchers need clear guidance about ELSI issues, so they can more readily ensure the responsible development and deployment of portable neuroimaging technologies.
2	Seven of the 11 interviewees confirmed that their portable neuroimaging technology will often rely on AI, cloud-computing, and cloud data storage to achieve cost-effective portability aims.	ELSI guidance needs to address the use of ethical and trustworthy artificial intelligence (ETAI). Portable neuroimaging technologies are being designed not just as stand-alone devices, but rather as part of a “stacked” set of technologies that regularly rely on AI.
3	Nine of the 11 interviewees are designing neuroimaging technologies that can be deployed far from brick-and-mortar healthcare sites.	ELSI guidance must take into consideration that portable neuroimaging devices are being engineered for researchers and clinicians to use in remote locations. These technologies raise concerns about barriers to clinical evaluation of incidental findings, safety of device setup in remote locations, and the capacity and training that those deploying this technology have to interpret and utilize the information generated by the device.
4	The interviews revealed that the salience of specific ELSI issues varies by technology. For example, identifying and planning a pathway to evaluate incidental findings was of high importance to developers of portable MRI, but not to developers of HD-DOT and portable EEG.	Although some general ethical principles will apply across all modalities of portable neuroimaging, ELSI guidance will need to address issues specific to different imaging modalities.

3.1 |. Finding 1: The ELSI guidance gap

The interviews revealed that developers are unaware of ELSI guidance for research with portable neuroimaging technologies. There is a need to generate clear guidance and awareness within the scientific and engineering communities that are building and deploying portable neuroimaging machines. One of the issues we probed with each interviewee was whether their teams were aware of, or had applied, ELSI guidance for development or deployment of portable neuroimaging technologies. None of the 11 interview participants were aware of such guidance. One researcher who has conducted field-based fNIRS research was not aware of “any sort of established ethical guide” and observed that IRB guidance was overly “general” and not specific to the issues emerging in fNIRS research (Interview 6). Another researcher, working on pMRI systems, noted that ethical issues were “not discussed in the grant” that was funding the technology development (Interview 11).

Moreover, multiple interviewees reported that their research teams had given little consideration to ELSI issues. For instance, a researcher who conducts optical neuroimaging research in the field described a research study in which participants could potentially receive a copy of their brain scan conducted as part of the research study. When probed about whether a strategy had been created to address the ethical concern that participants might misunderstand the meaning and utility of the returned brain scan, the respondent replied that such an ethics strategy “was not built into the study design” (Interview 1). Another interviewee developing pMRI technology was asked if there were ever conversations about ethical issues within the research team, to which they replied: “No. Not at all” (Interview 4).

The finding that there is a lack of ELSI guidance for portable neuroimaging developers and researchers is consistent with conclusions of the Working Group we led that focused on pMRI research. As described above, the WG engaged in a multi-year process of analysis that identified “key gaps in current ethical and legal guidance applicable to portable MRI.”¹ In addition, this finding from the interviews is consistent with results from a survey of the general public about pMRI, which found that members of the public generally did not rate most potential ELSI concerns with pMRI research as important.³⁷ When we separately surveyed expert stakeholders, including experts in ethics and law, we found that those experts did identify a number of potential ELSI concerns with pMRI research as important.³⁸ Taken together, these multiple data points suggest that there is a need to better inform both the general public and the scientific community about potential ethical issues and how to address them.

3.2 |. Finding 2: “Stacked” technologies will utilize AI and cloud computing

Because many portable neuroimaging technologies are being designed as part of a “stacked” set of technologies that regularly incorporate AI, ELSI guidance needs to integrate guidance on ethical and trustworthy artificial intelligence (ETAI). AI has been recognized in the literature as a key factor facilitating the rise of tele-neurology, in which neurological care is provided across great distances.⁴¹ The use of AI in portable neuroimaging may also reduce costs and save space by reducing the need for some larger and more expensive hardware.⁴² Seven of the 11 interviewees confirmed that their portable neuroimaging technology will often rely on AI, cloud-computing, and cloud data storage to achieve cost-effective portability aims. One MRI developer stated they “have aspirations of putting in deep learning and AI based [image] reconstruction” (Interview 9). An EEG researcher working with wearable devices said their project has a team of people “trying to train their AI” to accurately delineate seizure activity from raw EEG data (Interview 2). An MRI developer stated that AI will increase the utility of their images by “denoising the images,” resulting in a product that “a radiologist is going to have more confidence in” (Interview 9). Likewise, another interviewee said that portable EEG innovators are counting on “the artificial intelligence that [they] have trained to recognize this seizure pattern” and orchestrate automatic, neuro-stimulation interventions outside of the clinic setting (Interview 2). Although there is a well-developed literature on the ethics of AI in medicine and health care,^43,44 ethical concerns about AI were voiced by only a few of the interviewees.

Cloud-based MRI analysis utilizing AI is being coupled with user-friendly designs that envision a pMRI machine capable of being operated by someone with very little technical proficiency, and in the future, self-service neuroimaging may require no operator at all.⁴⁵ As we have explored in previous work, the use of cloud data storage and cloud-computing is also a critical component to enable portability.^1,17 The reduced MRI footprint may allow a second MRI developer to do scanning in dynamic scenarios like “disaster rescue” (Interview 3). Disaster rescue refers to the phase of immediate response after a disaster, “including triage and medical care for disaster victims.”⁴⁶ MRI imaging has historically not been possible on-site to inform triage decisions, but this interviewee anticipates this future possibility.

3.3 |. Finding 3: Expanding access and ease of use

In contrast to most traditional fixed neuroimaging devices, which must be operated by a trained expert, the interviews revealed that portable neuroimaging devices are being engineered for utilization even by those who do not have extensive experience in neuroimaging. While the technology may expand MRI access by allowing a research team with no prior experience to acquire brain data at the push of a button, concerns arise if the research team does not have sufficient expertise to appropriately interpret the brain data.

One developer of pMRI plans “to make it [have a] small footprint, compact” so that brain imaging will “be available to a lot of places, especially … those places that [don’t] have a stable setup” (Interview 3). A second developer of pMRI posits that as the devices become increasingly portable, researchers “can start to get into some really naturalistic environments” (Interview 6). While these devices could operate in traditional settings, our interviews found that portable neuroimaging technology will allow for much more remote data collection. Nine of the 11 interviewees are designing neurotechnology that can be deployed far from brick-and-mortar healthcare sites.

These devices are designed to require minimal expertise to run on-site. One HD-DOT developer stated that the technology was “certainly easy enough for an undergraduate student to pick up” (Interview 1). A developer of pMRI said that in order to maximize mobility, they “can’t expect people who would operate it [to] have … expertise, like [in] a proper hospital” (Interview 3). A second developer of pMRI stated that the goal is to make the device operation “simple” so that “somebody relatively untrained would literally just push a button” (Interview 11). A third developer of pMRI aims to create a device that “will be very easy to use and intuitive so that anyone can use it” (Interview 9).

Increased ease of use should greatly expand the number of research and clinical teams that can utilize portable neuroimaging. The developments described above suggest that brain data will soon be able to be acquired by someone with quite limited neuroimaging expertise. But as we have discussed in more detail elsewhere,^1,47 a large increase in the number of brain scans acquired in remote and under-resourced communities will require research teams to put into place research protocols that ensure sufficient expertise on site to obtain informed consent and answer participant questions, expertise on the team to interpret and explain brain scans to participants, and strategies to “ensure pathways to timely care in the event of incidental findings or concerning research results, regardless of the participant’s geographic location, insurance status, and ability to pay for care.”¹

3.4 |. Finding 4: Ethical concerns vary by imaging modality

Although some ELSI guidance may be generalizable across these different portable neuroimaging technologies, such as the AI issues discussed above, the interview data show that the salience of ELSI issues will vary by modality. This suggests the need for tailored guidance.

The issue of incidental findings illustrates the need for technology-specific guidance. For pMRI research generating structural brain scans, identifying and planning a pathway for access to further evaluation of incidental findings will be of high importance.⁴⁷ However, developers of other technologies, such as HD-DOT and portable EEG, were much less concerned. As one such developer said, “I would be impressed, baffled, and shocked if we were capable of getting an incidental finding [using HD-DOT]” (Interview 1). Another noted that “incidental findings is less of a concern now with EEG” (Interview 2).

The risk of re-identification based on structural brain images was not of concern for those interviewees who are developing fNIRS and EEG systems. When asked about the possibility of re-identification with only fNIRS (and no structural) data, an fNIRS researcher responded: “I don’t think there is a concern there” (Interview 1). Similarly, an EEG researcher observed that unless someone conducted highly technical EEG data collection “of every human” in the world, re-identification via EEG data alone is not an issue as a practical matter (Interview 5).

4 |. ELSI GUIDANCE FOR PORTABLE NEUROIMAGING

This exploratory study of the views held by those developing and implementing new portable neuroimaging technologies confirms that deployment of these technologies for research raises a host of ELSI issues. Yet the respondents in our sample were generally unaware of, or not deeply engaged with, these ethical concerns. For example, a researcher conducting real-time EEG research in the field stated that they had not addressed privacy concerns yet in their work and were “still … in a fog about what [type of privacy protection] is actually required”(Interview 5).

This is not to say that all interviewees were unaware of ELSI issues. Several interviewees called out ELSI issues that had drawn their attention. For instance, while discussing ethical issues posed by conducting neuroimaging research in under-resourced countries, one researcher indicated that the desire to avoid the therapeutic misconception by dispelling participant expectations that scanning for research would provide medical treatment “was more [their] personal bias or personal preference” (Interview 1). The same researcher was concerned with how to “constrain people’s interest in overinterpreting functional neuroimaging” and manage participant expectations when the study was “one of their very few, very limited interactions with the health care system” (Interview 1).

As noted above, to deepen the engagement of neuroscientists and technology developers on ethical issues, the National Institutes of Health (NIH) has funded a series of projects (3U01EB025153–02S2; U01EB025153–03S1; and RF1MH123698) in which the senior authors on this paper (F.L., F.S., S.W.) have been working with multidisciplinary colleagues to advance that work.^1,23 The BRAIN Initiative has also convened ethics experts to issue guidance on a range of ELSI issues in neuroscience,^48–50 though not specifically on highly portable neuroimaging. Meanwhile, professional societies are now contributing to ELSI work; for example, in its inaugural workshop on low-field MRI in 2022, the International Society for Magnetic Resonance in Medicine (ISMRM) integrated an ELSI presentation.⁵¹ Our interviews support the need for progress on the ELSI issues raised by highly portable neuroimaging.

Of particular relevance for the theme of this special issue, MRI for the World, is guidance published by two of the authors on this paper (F.S. and S.W.), in collaboration with a “multi-disciplinary Working Group including 10 scientists with experience conducting neuroscience research in remote or resource-poor communities ….”² In that published guidance, we identified two overarching principles and five core ELSI issues that “should be addressed in developing guidance for field-based MRI research studies in resource-limited contexts.” We provide a high-level summary of that guidance in Figure 1; readers can review the full article for detailed discussion of these recommendations.² The guidance was developed specifically for pMRI, though many of the recommendations would be applicable to other portable neuroimaging modalities as well.

FIGURE 1.

Core ELSI issues, with key Working Group recommendations for addressing them, in the life cycle of field-based MRI research in remote and resource-limited communities. This figure summarizes guidance provided in Shen et al.² The figure is reproduced with permission of NeuroImage.

For the pMRI research community, our NIH BRAIN Neuroethics project offers additional resources and recommendations to address gaps in ELSI guidance identified by interviewees. From 2020–2024, a Working Group hosted by the University of Minnesota Consortium on Law and Values in Health, Environment & the Life Sciences developed 15 core pMRI research ELSI issues, guidance on those issues, and recommended solutions.¹ Of particular importance for pMRI developers and those who will be deploying pMRI in research are these recommendations (each discussed at much greater length in Shen et al.¹):

• “Each member of the portable MRI research team should have demonstrated competence to carry out their research role. For instance, scanner operators should have demonstrated competence to safely operate the portable MRI machine.” Just as we require a license to drive a car, so too must there be training requirements for operation of a portable neuroimaging machine. This would apply as well to do-it-yourself MRI machines.

• “Before carrying out the research, researchers conducting portable MRI research should become familiar with … ELSI issues … and investigators designing research should partner with the local communities in which research will occur.” When we conducted our interviews, ethical guidance for pMRI had not yet been published. But as this guidance is now starting to emerge, it is incumbent upon developers and researchers to utilize the guidance.

• “Safety guidelines and education should be created by the ACR, ISMRM, and other professional bodies, for use of highly portable MRI in field settings. These guidelines should cover safe setup, use, storage, and transport of the equipment and standards for participant privacy and data security.” The creation of updated guidelines by relevant professional bodies would address the gaps that our interviewees identified, and serve as an important guard against unsafe deployment of portable neuroimaging.

To facilitate implementation of our project’s recommendations, we developed a Portable MRI Research ELSI Checklist.²³ The checklist provides researchers, including those who are new to ELSI issues, with practical operational guidance for carrying out pMRI research that adheres to best ELSI practices. Given our finding in these interviews that ELSI issues vary by modality, the pMRI checklist would need modality-specific modification for some items.

5 |. CONCLUSION

Our interviewees anticipate that an array of portable neuroimaging devices will allow experts and non-experts alike to conduct brain research in previously inaccessible locations and with previously excluded populations. This future will rely heavily on “stacked” technologies that incorporate AI and cloud computing. But our interviews also reveal a critical gap: a lack of engagement with ELSI issues that will accompany these technological advances and a lack of available ELSI guidance at the development and early research stage. Our interviews suggest that many developers have not yet had an opportunity to deeply engage with neuroethics. But when we raised the issues, they recognized the importance of addressing ethical challenges that may arise in their work. This exploratory study supports the need to involve developers in consideration of the issues raised by these new highly portable technologies and confirms the need to develop ethical guidance for use of these technologies in field-based neuroimaging research as portable neuroimaging for the world emerges. Such guidance is becoming available through the publication of consensus guidelines for ethical research with pMRI and with the publication of the Portable MRI Research ELSI Checklist.²³ This guidance can help ensure the safe deployment of the emerging portable technologies that can revolutionize brain research across the globe.

Supplementary Material

SUPPORTING INFORMATION

Additional supporting information can be found online in the Supporting Information section at the end of this article.

Abbreviations

AI

artificial intelligence

BRAIN

Brain Research through Advancing Innovative Neurotechnologies Initiative

CESTI

Committee on Emerging Science, Technology, and Innovation

DOE

US Department of Energy

DTC

direct-to-consumer

EEG

electroencephalography

ELSI

ethical, legal, and societal issues

ETAI

ethical and trustworthy artificial intelligence

FDA

US Food and Drug Administration

fNIRS

functional near-infrared spectroscopy

HD-DOT

high-density diffuse optical tomography

HIPAA

Health Insurance Portability and Accountability Act

ICU

intensive care unit

IRB

Institutional Review Board

ISMRM

International Society for Magnetic Resonance in Medicine

MEG

magnetoencephalography

MRI

magnetic resonance imaging

NASEM

National Academies of Sciences, Engineering, and Medicine

NIH

National Institutes of Health

PET

positron emission tomography

pMRI

portable MRI

DATA AVAILABILITY STATEMENT

Requests for access to the data that support the findings of this study will be considered as received.