News & Comment

Cybersecurity is an increasingly important concern for reliable healthcare delivery and is particularly salient for robotic surgery. Surgical robots are complex systems with numerous points of vulnerability, and there have been real-world demonstrations of successful cyberattacks on surgical robots. There are several ways to improve the risk profile of robotic surgery, including recognizing system complexity, investing in regular software updates, following cybersecurity best-practices, and increasing transparency for all stakeholders. As robotic surgery continues to technologically advance, ensuring overall system safety from a cybersecurity perspective is paramount.

In the metaverse, users will actively engage with 3D content using extended reality (XR). Such XR platforms can stimulate a revolution in health communication, moving from information-based to experience-based content. We outline three major application domains and describe how the XR affordances (presence, agency and embodiment) can improve healthy behaviour by targeting the users’ threat and coping appraisal. We discuss how health communication via XR can help to address long-standing health challenges.

Artificial intelligence (AI) tools for endoscopy are now entering clinical practice after demonstrating substantial improvements to polyp detection on colonoscopy. As this technology continues to mature, efforts to develop and validate a new frontier of possibilities—including diagnostic classification, risk stratification, and clinical outcomes assessment—are now underway. In npj Digital Medicine, scientists from Cosmo AI/Linkverse and collaborators report an extension to the first FDA-cleared AI tool for colonoscopy that goes beyond polyp detection to enable video-based diagnostic characterization.

The text-guided diffusion model GLIDE (Guided Language to Image Diffusion for Generation and Editing) is the state of the art in text-to-image generative artificial intelligence (AI). GLIDE has rich representations, but medical applications of this model have not been systematically explored. If GLIDE had useful medical knowledge, it could be used for medical image analysis tasks, a domain in which AI systems are still highly engineered towards a single use-case. Here we show that the publicly available GLIDE model has reasonably strong representations of key topics in cancer research and oncology, in particular the general style of histopathology images and multiple facets of diseases, pathological processes and laboratory assays. However, GLIDE seems to lack useful representations of the style and content of radiology data. Our findings demonstrate that domain-agnostic generative AI models can learn relevant medical concepts without explicit training. Thus, GLIDE and similar models might be useful for medical image processing tasks in the future - particularly with additional domain-specific fine-tuning.

Responsible adoption of healthcare artificial intelligence (AI) requires that AI systems which benefit patients and populations, including autonomous AI systems, are incentivized financially at a consistent and sustainable level. We present a framework for analytically determining value and cost of each unique AI service. The framework’s processes involve affected stakeholders, including patients, providers, legislators, payors, and AI creators, in order to find an optimum balance among ethics, workflow, cost, and value as identified by each of these stakeholders. We use a real world, completed, an example of a specific autonomous AI service, to show how multiple “guardrails” for the AI system implementation enforce ethical principles. It can guide the development of sustainable reimbursement for future AI services, ensuring the quality of care, healthcare equity, and mitigation of potential bias, and thereby contribute to realize the potential of AI to improve clinical outcomes for patients and populations, improve access, remove disparities, and reduce cost.

Over the past 7 years, regulatory agencies have approved hundreds of artificial intelligence (AI) devices for clinical use. In late 2020, payers began reimbursing clinicians and health systems for each use of select image-based AI devices. The experience with traditional medical devices has shown that per-use reimbursement may result in the overuse use of AI. We review current models of paying for AI in medicine and describe five alternative and complementary reimbursement approaches, including incentivizing outcomes instead of volume, utilizing advance market commitments and time-limited reimbursements for new AI applications, and rewarding interoperability and bias mitigation. As AI rapidly integrates into routine healthcare, careful design of payment for AI is essential for improving patient outcomes while maximizing cost-effectiveness and equity.

Healthcare is a large contributor to greenhouse gas (GHG) emissions around the world, given current power generation mix. Telemedicine, with its reduced travel for providers and patients, has been proposed to reduce emissions. Artificial intelligence (AI), and especially autonomous AI, where the medical decision is made without human oversight, has the potential to further reduce healthcare GHG emissions, but concerns have also been expressed about GHG emissions from digital technology, and AI training and inference. In a real-world example, we compared the marginal GHG contribution of an encounter performed by an autonomous AI to that of an in-person specialist encounter. Results show that an 80% reduction may be achievable, and we conclude that autonomous AI has the potential to reduce healthcare GHG emissions.

Due to its enormous capacity for benefit, harm, and cost, health care is among the most tightly regulated industries in the world. But with the rise of smartphones, an explosion of direct-to-consumer mobile health applications has challenged the role of centralized gatekeepers. As interest in health apps continue to climb, national regulatory bodies have turned their attention toward strategies to protect consumers from apps that mine and sell health data, recommend unsafe practices, or simply do not work as advertised. To characterize the current state and outlook of these efforts, Essén and colleagues map the nascent landscape of national health app policies and raise several considerations for cross-border collaboration. Strategies to increase transparency, organize app marketplaces, and monitor existing apps are needed to ensure that the global wave of new digital health tools fulfills its promise to improve health at scale.

As clinicians and scientists gather more data on the clinical trajectory of COVID-19 and the biology of its causative agent, the SARS-CoV-2 virus, novel strategies are needed to integrate these data to inform new therapies. A recent study by Howell et al. introduces a network model of viral-host interactions to produce explainable and testable predictions for treatment effects. Their model was consistent with experimental data and recommended treatments, and one of its predicted drug combinations was validated through in vitro assays. These findings support the utility of computational strategies for leveraging the vast literature on COVID-19 to generate insights for drug repurposing.

Health care is a human process that generates data from human lives, as well as the care they receive. Machine learning has worked in health to bring new technology into this sociotechnical environment, using data to support a vision of healthier living for everyone. Interdisciplinary fields of research like machine learning for health bring different values and judgements together, requiring that those value choices be deliberate and measured. More than just abstract ideas, our values are the basis upon which we choose our research topics, set up research collaborations, execute our research methodologies, make assessments of scientific and technical correctness, proceed to product development, and finally operationalize deployments and describe policy. For machine learning to achieve its aims of supporting healthier living while minimizing harm, we believe that a deeper introspection of our field’s values and contentions is overdue. In this perspective, we highlight notable areas in need of attention within the field. We believe deliberate and informed introspection will lead our community to renewed opportunities for understanding disease, new partnerships with clinicians and patients, and allow us to better support people and communities to live healthier, dignified lives.

Continued COVID-19 surges have highlighted the need for widespread testing in addition to vaccination for disease containment. SARS-COV-2 RNA can be found in faecal matter, making human stool another potential source for COVID-19 diagnostics. In this commentary, we highlight potential strategies to use a smart toilet platform to passively monitor COVID-19 surges, enabling earlier detection of infected individuals and promoting public health.

Increasing digitization across the healthcare continuum has revolutionized medical research, diagnostics, and therapeutics. This digitization has led to rapid advancements in the development and adoption of Digital Health Technologies (DHT) by the healthcare ecosystem. With the proliferation of DHTs, the term ‘digital biomarker’ has been increasingly used to describe a broad array of measurements. Our objectives are to align the meaning of ‘digital biomarker’ with established biomarker terminology and to highlight opportunities to enable consistency in evidence generation and evaluation, improving the assessment of scientific evidence for future digital biomarkers.

The vital signs—temperature, heart rate, respiratory rate, and blood pressure—are indispensable in clinical decision-making. These metrics are widely used to identify physiologic decline and prompt investigation or intervention. Vital sign monitoring is particularly important in acute care settings, where patients are at higher risk and may require additional vigilance. Conventional contact-based devices, while widespread and generally reliable, can be inconvenient or disruptive to patients, families, and staff. Non-contact, video-based methods present a more flexible and information-dense alternative that may enable creative improvements to patient care. Still, these approaches are susceptible to several sources of bias and require rigorous clinical validation. A recent study by Jorge et al. demonstrates that video-based monitoring can reliably capture heart rate and respiratory rate and overcome many potential sources of bias in post-operative settings. This presents real-world evaluation of a practical, noninvasive, and continuous monitoring technology that had previously only been tested in controlled settings.