JMIR Publications

ABSTRACT

Background:

Cardiac rehabilitation is a core measure for secondary prevention of cardiovascular disease, but global participation rates remain consistently low. Robotics, with its precise movement assistance capabilities and personalized interactive functions, offers a new approach to overcoming the limitations of traditional rehabilitation models. However, research in this field is diverse in design and evidence remains scattered, and there is a lack of systematic review and integration.

Objective:

This study aims to use a scoping review methodology to systematically review the current status of robotics applications in cardiac rehabilitation for patients with cardiovascular disease. It seeks to summarize the types of technologies, target populations, rehabilitation stages, and outcome measures; clarify the scope and distribution characteristics of existing evidence; identify knowledge gaps; and provide a reference basis for future research design and clinical practice.

Methods:

We conducted this review from March to July 2026. With one lead researcher and two professional librarians, we developed search strategies and systematically searched three Chinese databases (CNKI, Wanfang, VIP) and four English databases (PubMed, Web of Science, Cochrane Library, Embase) from inception to April 4, 2026. Two independent reviewers screened studies, extracted key data, and resolved disagreements by consulting two additional reviewers.

Results:

This study ultimately included 10 articles involving 587 patients with cardiovascular disease from 5 countries. The published articles spanned the years 2015 to 2025, and the study designs included 3 randomized controlled trials and 7 non-randomized studies. Robotic technologies can be categorized into two main types: physical rehabilitation robots—which encompass four technical architectures, including exoskeleton gait training, balance training, motorized resistance training, and soft wearable devices, accounted for 70% of the included studies; social assistance robots, with voice interaction and personalized feedback as their core functions, accounted for 30%. In terms of rehabilitation phases, 7 studies focused on Phase II outpatient rehabilitation, 3 involved Phase I inpatient rehabilitation, and there were no studies on the Phase III community maintenance phase. Outcome measures covered four dimensions: physiological, functional, psychosocial, and safety. Robot-assisted training demonstrated positive effects in improving cardiopulmonary function, enhancing motor abilities, and increasing patient engagement, and its safety was preliminarily validated across populations with varying disease severities.

Conclusions:

Robotic technology is driving the transformation of cardiac rehabilitation from a traditional model dominated by human supervision to an intelligent model of human-robot collaboration. The two types of robots complement each other functionally, collectively addressing the core challenges of traditional rehabilitation regarding individualized supervision and maintaining motivation among patients with cardiovascular disease. This review provides a conceptual framework for future research and helps advance the field from fragmented exploration toward systematic and standardized development.