Robot-assisted upper-limb rehabilitation has been studied for many years, with many randomised controlled trials (RCTs) investigating the effects of robotic-assisted training on affected limbs. The current trend directs towards end-effector devices. However, most studies have focused on the effectiveness of rehabilitation devices, but studies on device sizes are relatively few. Systematically review the effect of a portable rehabilitation robot (PRR) on the rehabilitation effectiveness of paralysed upper limbs compared with non-robotic therapy. A meta-analysis was conducted on literature that included the Fugl-Meyer Assessment (FMA) obtained from the PubMed and Web of Science (WoS) electronic databases until June 2023. A total of 9 studies, which included RCTs, were completed and a meta-analysis was conducted on 8 of them. The analysis involved 295 patients. The influence on upper-limb function before and after treatment in a clinical environment is analysed by comparing the experimental group using the portable upper-limb rehabilitation robot with the control group using conventional therapy. The result shows that portable robots prove to be effective (FMA: SMD = 0.696, 95% = 0.099 to.293, p < 0.05). Both robot-assisted and conventional rehabilitation effects are comparable. In some studies, PRR performs better than conventional rehabilitation, but conventional treatments are still irreplaceable. Smaller size with better portability has its advantages, and portable upper-limb rehabilitation robots are feasible in clinical rehabilitation. Although portable upper-limb rehabilitation robots are clinically beneficial, few studies have focused on portability. Further research should focus on modular design so that rehabilitation robots can be decomposed, which benefits remote rehabilitation and household applications.

The use of virtual reality (VR) for stroke rehabilitation has been implemented during the last decade. At present, most studies still focus on the effects of VR on upper limb rehabilitation, and few studies have explored VR games, VR system designs, and rehabilitation modes for upper limb rehabilitation. This study aims to (1) evaluate the rehabilitation effect of stroke patients using a game-based VR upper limb rehabilitation system in clinical settings; (2) investigate the impact of custom and commercial VR games on patients in clinical settings; and (3) review VR upper limb rehabilitation modes. The PubMed, ScienceDirect, Scopus, Web of Science, and IEEE Xplore databases were searched, and related literature published through December 2021 was included. A total of 4700 articles were retrieved according to the search strategy. We identified 24 studies, including 793 patients. We conducted a systematic search for randomized controlled trials with adult stroke patients to analyze the effect of game-based VR upper limb rehabilitation systems. A meta-analysis was conducted to compare the effects of upper limb function, hand dexterity, daily living ability, and cognitive function between the experimental group (EG, using VR for upper limb rehabilitation) and control group (CG, receiving conventional rehabilitation, including physical therapy and occupational therapy). We also conducted an analysis of both custom and commercial games. The results of the meta-analysis proved that game-based VR upper limb rehabilitation therapy for cerebral apoplexy is an effective method of rehabilitation in clinical settings and is more effective than traditional rehabilitation in improving patients' upper limb function and hand mobility. Custom games heal better than commercial games. This study only includes nonimmersive device rehabilitation modes due to research constraints and classified them into four categories. The mode of VR games combined with rehabilitation instruments may solve the problem that patients with severe upper limb dysfunction cannot operate games. Whether the use of immersive VR devices and the fun of games will affect patients' rehabilitation motivation and effect is the direction of future research.

Study design: case series. Introduction: Robot-assisted therapy for upper limb rehabilitation is an emerging research topic and its design process must integrate engineering, neurological pathophysiology and clinical needs. Purpose of the Study: This study developed/evaluated the usefulness of a novel rehabilitation device, the MirrorPath, designed for the upper limb rehabilitation of patients with hemiplegic stroke. Methods: The process follows Tseng’s methodology for innovative product design and development, namely two stages, device development and usability assessment. During the development process, the design was guided by patients’ rehabilitation needs as defined by patients and their therapists. The design applied synchronic movement of the bilateral upper limbs, an approach that is compatible with the bilateral movement therapy and proprioceptive neuromuscular facilitation theories. MirrorPath consists of a robotic device that guides upper limb movement linked to a control module containing software controlling the robotic movement. Results: Five healthy subjects were recruited in the pre-test, and 4 patients, 4 caregivers and 4 therapists were recruited in the formal test for usability. All recruited subjects were allocated to the test group, completed the evaluation and their data were all analyzed. The total system usability scale (SUS) score obtained from the patients, caregivers and therapists was 71.8±11.9, indicating a high level of usability and product acceptance. Discussion and Conclusions: Following a standard development process, we could yield a design that meets clinical needs. This low-cost device provides a feasible platform for carrying out robot-assisted bilateral movement therapy of patients with hemiplegic stroke. (clinical trial registration number NCT02698605)