Embodied intelligence, involving artificial intelligence integrated into physical or virtual bodies for human interaction, is gaining traction in healthcare. Despite its increasing application, the evidence base remains fragmented due to inconsistent terminology, varied embodiment forms, and limited synthesis of application domains, target populations, care settings, acceptability, and effectiveness. This hampers clarity and translation into daily healthcare practices.
To address these gaps, a scoping review was conducted to comprehensively map the applications of embodied intelligence in healthcare. The review systematically classified five embodiment forms, identified three major functional domains, described target populations and implementation settings, and synthesized evidence on acceptability and effectiveness.
The findings from the review revealed that virtual humanoid agents and physical humanoid robots were the most utilized forms of embodied intelligence in health care. Applications were concentrated in health management and health education, as well as mental health promotion. Older adults were the primary target population, with interventions commonly implemented in home settings, care homes, labs, and hospitals. Notably, randomized controlled trials reported positive effects on health behaviors, mental health outcomes, and cognitive function.
The review emphasized the importance of consolidating fragmented evidence and standardizing terminology to facilitate the implementation of embodied intelligence in healthcare. It called for standardized outcome measures, rigorous trials, and longitudinal evaluations to enable scalable and ethically grounded real-world adoption.
Embodied intelligence has shown promise in healthcare applications, aiding in health management, education, mental health support, and physiological health promotion. The review highlighted gaps in the evidence base, such as limited representation of certain embodiment forms and domains, as well as the need for robust research methodologies to support the integration of embodied intelligence into routine healthcare practices. Future studies are encouraged to focus on ecologically valid evaluations and standardized reporting for meaningful translation of embodied intelligence into patient-centered care.
