Editorial: Machine Learning and Immersive Technologies for User-Centered Digital Healthcare Innovation

Federico Colecchia ^1* Daniele Giunchi ² Rui Qin ³ Eleonora Ceccaldi ⁴ Fang Wang ¹

• ¹ Brunel University London, Uxbridge, United Kingdom
• ² University College London, London, England, United Kingdom
• ³ University of Leicester, Leicester, East Midlands, United Kingdom
• ⁴ University of Genoa, Genoa, Liguria, Italy

Whereas the Research Topic title explicitly refers to healthcare innovation, the broader scope has turned this article collection into an opportunity for reflection on a range of topics of relevance to both health and wellbeing. Topics have included machine learning and immersive technologies for enhancing the provision of medical education and training, for improving workforce wellbeing, and for augmenting art therapy programmes with a view to increasing therapeutic compliance. Methods employed include Agile data science techniques, 'CRoss Industry Standard Process for Data Mining' (CRISP-SM), 'Preferred Reporting Items for Systematic reviews and Meta-Analyses' (PRISMA), thematic literature review, mini review, primary research (specifically, collection of data from university nursing students and surgeons), 'Simulation Effectiveness Tool -Modified' (SET-M), established data science techniques, and human factors engineering methods. Key research themes that have emerged from the Research Topic are discussed below, followed by a reflection on priorities for further research and development.The articles have highlighted a need for knowledge and expertise from across academic disciplines and professional practice to converge and underpin the development of digital innovations promoting individuals' health and wellbeing. Relevant disciplines and domains include computer science, user-centred design, human-computer interaction, engineering, human factors engineering, and the social sciences. Technology end user and stakeholder values, expectations, and requirements need to be addressed if new methods enabled by modern digital technologies are to be sustainably employed [1][2][3][4]. Interestingly, the articles have generally suggested the importance of embedding end user and stakeholder perspectives within technology development workflows, although only a minority of the studies have explicitly articulated the need for extensive involvement of humancentred design researchers and practitioners for scaffolding and facilitating iterative development and evaluation [3]. Unsurprisingly, the need to address ethical concerns appears intertwined with the recognised desirability of research objectives and methods to deliver deeper integration across discipline boundaries. This is illustrated by research advocating the adoption of intersectional social sciences perspectives within AI development for cancer diagnostics [2] and by studies focussing on the representativity of machine learning training data to reduce health inequalities affecting specific ethnic groups in relation the provision of diagnostic services [3].The design and development of production-ready AI-based systems designed for flexibility and maintainability over time have received significant attention in recent years, particularly in the information systems, human-computer interaction, and engineering design literature. This is illustrated by Research Topic articles focussing on the definition of architectural requirements for healthcare cost estimation systems relying on dedicated predictive numerical models [5], and by studies delivering prototype models to enable healthcare professionals to query different Electronic Medical Record systems using intuitive interfaces based on natural language [6]. Such efforts have achieved a balance between adapting research pipelines for production environments and identifying optimised architectural specifications from an information systems perspective.The emphasis in recent academic and professional discourse on opportunities afforded by immersive technologies, including virtual reality, augmented reality, and mixed reality, for achieving more efficient and inclusive delivery of medical educational and training programmes is reflected in this article collection. An interesting review study has focussed on a comparison between technologyaugmented methods and established approaches [7]. Reported benefits include enhanced student and trainee motivation, satisfaction, and learning outcomes, although the possible occurrence of undesired consequences of the use of immersive technologies, including cybersickness following prolonged exposure, has been noted. Interestingly, one study has focussed on real-time detection of cybersickness with a view to reducing detrimental effects on user experience [8]. Proposed innovations based on mixed reality to streamline urology anatomy training and to facilitate preoperative urology planning have attracted positive feedback from both university nursing students and surgeons, which encourages further research towards more extensive clinical validation [9]. An interesting study has focussed on an integration between generative AI and immersive technologies for designing augmented reality filters, with a view to improving medical students' perceptions of self-efficacy in recognising selected disease manifestations [10].The potential of modern digital technologies for improving individuals' wellbeing has been the subject of recent research, which is reflected in this article collection. The breadth of contributions received illustrates the potential of artificial intelligence and immersive technologies for improving individuals' wellbeing, particularly in clinical and workplace settings. A theoretical model has been presented, explaining the psychological benefits of virtual immersion for oncology patients with emphasis on distraction for alleviating anxiety and pain [4]. Opportunities have been identified for artistic expression within virtual reality environments to increase therapeutic compliance and to improve wellbeing outcomes for individuals in relation to psychotherapy and neurorehabilitation [11]. A study has identified features of immersive technologies that hold potential for improving motor rehabilitation compliance and efficacy with stroke patients when used in combination with traditional approaches [12]. Such features include those enabling real-time movement tracking and the provision of reinforced feedback in line with established neurorehabilitation principles. A review of modern digital technologies for estimating individuals' wellbeing in workplace settings has generated useful recommendations on how real-time posture detection is best combined with the adoption of established human factors engineering best practices [13]. This has enabled the identification of optimised algorithms to be employed in conjunction with physiological sensing methods towards the design of healthier workplaces.Overall, the articles published under this Research Topic have highlighted the importance of conducting interdisciplinary research when tackling challenges at the intersection of technology development with human-centred design and human factors engineering. Integrative capabilities across academic discipline silos and research methodologies -with emphasis on modern design research and design professional practice -have identified as an important enabler of challenge-driven research and responsible innovation. Such insights are relevant to the United Nations Sustainable Development Goal (SDG) number 3 ('Good health and well-being') and more broadly [14]. A balanced distribution has been achieved in this collection of articles between applications of immersive technologies [4;7-9;10-12] and applications of machine learning and artificial intelligence [1-3;5-6;8]. The authors speculate that future research is likely to reflect a convergence of immersive technologies and artificial intelligence in relation to the promotion of individuals' health and wellbeing. If that is the case, it is anticipated that the emphasis will be on human-centred design, participatory design, and methods addressing ethical issues of privacy, transparency, equitability, and fairness. One potential area of convergence relates to the development of personalised immersive experiences designed for inclusivity. It is expected that reliance on human-centred and participatory design methods will prove useful for scaffolding iterative design with significant involvement of technology end users and stakeholders. This is illustrated by the article discussing the use of artistic experiences within virtual reality environments to increase therapeutic compliance and to improve wellbeing outcomes [11]. The articles have also highlighted several limitations with the technology state of the art, which calls for additional emphasis on interdisciplinary research, human-centred design, and inclusive design research moving forward. Such limitations include the following: digital access barriers and reduced digital literacy across user groups; the generally reduced availability of dedicated features for visually-impaired individuals --and for individuals with specific characteristics more broadly -compared with mainstream users; undesired effects from the use of head-mounted immersive displays --including cybersickness; the presence of different skill sets within interdisciplinary AI development teams, potentially reducing the benefits of Agile development. Moreover, although the articles published under this Research Topic have not focussed on this aspect, future development will also need to address elements of clinical validation of digital technologies in the context of the relevant regulatory frameworks.