This study consists of two parts: (1) the evaluation of the content validity index and the preliminary usage of a prototype home-based multimodal exergame using a cross-sectional study design, and (2) the evaluation of the feasibility of the home-based multimodal exergame for patients with chronic stroke using a prospective single-group study design.

The inclusion criteria were (1) diagnosis of ischemic or hemorrhagic stroke at least 6 months before enrollment; (2) age between 20 and 80 years; (3) normal cognitive function, as determined by the Mental State Examination T10 score [29,30]; and (4) ability to see a notebook monitor screen from a distance of at least 60 cm, with or without corrective lenses. The exclusion criteria were (1) severe spasticity (Modified Ashworth Scale ≥3), (2) neglect or impaired communication, and (3) any comorbidity or complication that precluded participation in the training program.

The study participants consisted of 2 groups involved in different phases of the study: expert reviewers in study part 1 and patients with stroke in study parts 1 and 2. The expert panel was recruited to evaluate the content validity of the prototype home-based multimodal exergame. Moreover, 5 experts with experience in game and web-application design, health innovation, stroke rehabilitation, neurocognition, and physical-cognitive training participated in the content validation process.

For the preliminary usability evaluation and feasibility study, demographic characteristics of the participants, including age, sex, time post onset, paretic side, and education level, were recorded. Patients with chronic stroke were recruited using convenience sampling through online and offline advertising to local communities, hospitals, rehabilitation centers, and social media advertisements. Participant screening and outcome assessments were conducted by blinded assessors.

As this study was designed as a preliminary usability and feasibility study, a formal power analysis was not performed. The intended and achieved sample sizes were 5 participants with chronic stroke for the preliminary usability evaluation and 9 participants for the feasibility study. Sample sizes were determined pragmatically based on feasibility and the exploratory nature of the study. No interim analyses or stopping rules were applied.

A novel home-based multimodal exergame was developed specifically for patients with chronic stroke by integrating the existing evidence from the literature [31,32] and adapted from our previous studies [33,34]. Additionally, a human-centered design methodology was applied to the development of the game-based exercise prototype [35], incorporating brainstorming sessions with the research team. Based on the synthesized evidence, a multimodal exergame-based intervention was applied to promote health performance, especially physical and cognitive aspects, among patients with stroke. The prototype of the multimodal exergame consists of physical exercises combined with cognitive training, including seated exercises, sit-to-stand training, proprioceptive neuromuscular facilitation (PNF) pattern-based exercises, balance training, aerobic exercise, and cognitive training targeting attention, memory, and executive function. Overall movements in multimodal exercise included shoulder flexion and extension, shoulder horizontal abduction and adduction, elbow flexion and extension, scapular protraction, lateral trunk flexion, hip flexion with knee flexion, hip abduction and adduction with knee flexion, knee flexion and extension, ankle dorsiflexion and plantarflexion. Gamification principles were integrated into the exergame to enhance user engagement, motivation, and adherence during exercise sessions. The overall structure and features of the game, including its objectives, gameplay mechanics, and graphical user interface, were systematically integrated and refined based on practical considerations to develop a functional prototype of the game-based training system. The exergame incorporated gamification features, including goal setting, rewards, interactive elements, and cognitive engagement, along with augmented feedback delivered through visual and auditory cues. Visual feedback (eg, on-screen movement tracking and progress bars) and auditory feedback (eg, sounds indicating success or errors) were integrated into the game-based design. Game-based concepts were applied to develop the digital exercise prototype using the Unity 3D game engine, incorporating computer vision techniques based on BlazePose for real-time pose estimation and user-system interaction. The system operates on a standard personal computer or notebook equipped with a webcam, without requiring additional wearable sensors. The system is semiautomated, enabling motion tracking and feedback without continuous therapist input. User-system interaction and feedback are achieved through real-time tracking of user movements, with feedback provided during gameplay as well as performance outcomes (eg, scores and errors) at the end of each game. The summary of the design and characteristics of home-based multimodal exergame is presented in Figure 1. Moreover, the novel home-based multimodal exergame includes a camera feature that displays participants’ movements, which can be monitored remotely through software such as Zoom (Zoom Communications) meetings. This functionality enables health care providers to remotely observe and assess the rehabilitation progress of patients with stroke at home or in remote locations, as illustrated in Figure 2.

The content validity of the developed intervention was assessed to determine whether its components accurately represent the intended construct for a specific clinical or research purpose. This evaluation was conducted by a panel of experts who rated each item based on its relevance and representativeness within the content domain. The content validity of the program was assessed by 5 experts who have expertise in the game and web-application design, innovation for health, stroke rehabilitation, neurocognition, and physical-cognitive training using the item-objective congruence (IOC) method. An overall IOC score greater than 0.75 for the intervention program was considered an acceptable content validity [36].

For the preliminary usability evaluation, written informed consent was obtained from all eligible participants before study participation. Demographic characteristics of participants, including age, sex, time post onset, paretic side, and education level, were recorded. A total of 5 patients with chronic stroke were enrolled to evaluate the preliminary usage of the game prototype. Participants meeting the eligibility criteria performed multimodal exergame. The multimodal exergame included a 10-minute warm-up and cool-down period, and 50 minutes of exercise. Participants were instructed to attend 2 sessions in a single visit to the research center. In session 1, participants were familiarized with the exergame via an in-game tutorial that provided detailed instructions on the correct performance of each exercise program. Following the tutorial, participants practiced a short sample of the exergame under the supervision of a physiotherapist. After the familiarization session, all participants were asked to rest until their perceived fatigue returned to baseline levels. In session 2, participants performed the multimodal exergame in the training room under the supervision of physiotherapists who remotely monitored them from the monitoring room. During exercise, participants were instructed to share their exergame desktop screen with the therapist via the Zoom meeting application. Additionally, the camera feature and computer vision-based system in the developed multimodal exergame allowed therapists to monitor exercise accuracy and safety.

After evaluating the content validity index and the preliminary usage of prototype home-based multimodal exergame, consensus from the expert and research team on the concept and design, along with comments and feedback, was used to refine the final version of the home-based multimodal exergame. The finalized version of the home-based multimodal exergame was used to evaluate the feasibility of the program in part 2 of the study.

For the evaluation of the feasibility of the home-based multimodal exergame, written informed consent was obtained from all eligible participants before participation in the study. This feasibility study was a prospective single-arm trial with blinded assessors conducted over a 6-week home-based intervention. A total of 9 patients with chronic stroke were recruited using convenience sampling to evaluate the feasibility of the exergame, applying the same inclusion and exclusion criteria as in part 1. Participants who met the eligibility criteria performed the intervention for 60 min/day, 3 days/week for 6 weeks. The home-based multimodal exergame is designed to be performed primarily in a seated position, with progression to functional tasks. The intervention comprises seated exercises, sit-to-stand training, PNF pattern-based exercises, balance training, aerobic exercise, and cognitive training. The multimodal exercise incorporates a range of joint movements and motor patterns involving the upper extremities, lower extremities, and trunk, with an emphasis on coordinated and functional movements such as alternating upper and lower limb movements, multidirectional reaching, weight shifting, trunk control, and sit-to-stand transitions. All exercises were performed for 5 repetitions per set for 2 sets, followed by 2 combined exercise sets (COMBO-SETs). The first COMBO-SET comprised alternating arm and leg lifts combined with forward arm and leg extensions, performed for 1 minute in level 1 and 2 minutes in level 2. The second COMBO-SET consisted of alternating boxing movements combined with alternating heel-toe tap steps, performed for 40 seconds in level 1 and 80 seconds in level 2. In addition, preparatory sit-to-stand training and sit-to-stand training were performed for 5 repetitions per set for 2 sets in level 1 and 3 sets in level 2. Cognitive tasks are embedded within the exergame to simultaneously engage attention, memory, and executive function. A summary of the home-based multimodal exergame is presented in Table 1. Before the home-based training, participants were asked to practice with a researcher to ensure they could perform the intervention correctly and safely. During the home-based intervention, participants were instructed to self-monitor their vital signs (eg, heart rate), rating of perceived exertion (RPE), and any adverse events. Participants were educated to recognize warning signs such as dizziness, loss of balance, or unusual symptoms, and were instructed to immediately stop exercising and notify the research team if such symptoms occurred. A researcher (PN) met with the participants via Zoom at the end of weeks 1, 2, and 4 to monitor their training. Each participant was assigned a unique Zoom ID, and all sessions were conducted on a one-to-one basis to ensure privacy and confidentiality. During the monitoring sessions, participants were instructed to share their screens with the researcher. In addition, the integrated camera function and computer vision-based system enabled real-time monitoring of exercise performance, allowing the researcher to assess movement accuracy and safety. The training sessions were recorded with participants’ consent. In addition, all recorded and collected data were anonymized and securely stored. Family members were permitted to stand beside participants during exercises at home.

For study part 1, usability outcomes and adverse events were collected immediately after the exergame session. For study part 2, physical and cognitive outcomes were assessed before and after the 6-week intervention. Program enjoyment was evaluated after 1 week and at the completion of the 6-week training period. Adherence and adverse events were also recorded throughout the study.

Due to the nature of the intervention, participants and researchers administering the intervention were aware of the study procedures. However, outcome assessments were conducted by blinded assessors (CS for cognitive assessment and PE for physical assessment) who were not involved in the intervention delivery.

All participants completed the study protocol and all outcome assessments, with no missing data or postdata collection exclusions.

All analyses and statistical calculations were performed using SPSS software (version 25.0, IBM Corporation). The content validity index of the program was determined using the IOC, with an IOC value greater than 0.75 indicating an acceptable level of validity. Descriptive statistics were used to describe the characteristics of participants and usability outcomes. The normality of the data was tested using the Shapiro-Wilk test. Paired sample t tests were used to compare the difference in outcome measures between pre- and post–6-week training. Mean differences were calculated, with 95% CIs reported. All data were presented as mean (SD) or n (%). A significant level was set at P<.05.

The study protocol was approved by the Ethics Committee of the Faculty of Associated Medical Sciences, Chiang Mai University (approval AMSEC-67EX-023). All participants provided written informed consent before participation in the study. All collected data were deidentified and handled in accordance with applicable ethical guidelines and data protection regulations to protect participant privacy and confidentiality. Participants received transportation reimbursement of 500 Thai Baht (THB; approximately US $15) in accordance with institutional research regulations. No identifiable personal information or participant images are included in this manuscript.

To the best of our knowledge, this intervention is the first prototype of an interactive, multimodal exercise program incorporating gamification and augmented feedback, specifically designed to target both cognitive and physical outcomes in individuals with chronic stroke residing at home or in remote areas. Findings from this study demonstrated that the preliminary use of the developed exergame prototype showed high content validity (IOC=0.82) and was enjoyable, easily accessible, user-friendly, and safe for patients with chronic stroke. Feasibility outcomes further support the implementation of this program, as reflected by a high adherence rate, absence of adverse events, and sustained enjoyment throughout the training period. Collectively, these findings suggest that the multimodal exergame is both safe and feasible for use in individuals with chronic stroke and may offer potential benefits for cognitive outcomes following 6 weeks of training.

Stroke rehabilitation interventions and poststroke exercise programs frequently encounter challenges, such as long travel distances, a lack of available health care providers, high dropout rates, and low motivation [13,14], all of which contribute to physical inactivity [11]. To mitigate rehabilitation barriers, this study developed and implemented a gamified physical-cognitive exercise program delivered through practical and accessible formats, including online videos and home-based training. Based on a synthesis of existing evidence [33,34,47,48] and a development brainstorming process, both physical and cognitive interventions for stroke rehabilitation were developed. All essential exercises for stroke rehabilitation were incorporated into the multimodal exercise program. Findings demonstrated that the content validity index of the overall program was high, indicating that the content of the home-based multimodal exergame is appropriate for individuals with chronic stroke. Additionally, interactive game-based training has emerged as a promising strategy to enhance user engagement, potentially leading to improved adherence and more favorable intervention outcomes [15,21,22]. The program in this study was designed according to gamification principles to promote user engagement and enjoyment. This study revealed that all participants reported high satisfaction and enjoyment with the program. The average score on PACES for each item was 5.97 ranged from 5.60 to 6.40, reflecting positive emotional responses and enjoyment during the combined physical and cognitive activities. The use of gamification elements may enhance enjoyment and promote long-term participation by aligning physical and cognitive challenges with game-like experiences [34,49,50]. These findings align with previous research demonstrating that gamification elements, such as goals, rules, and rewards, improved intrinsic motivation, satisfaction, and enjoyment in exercise programs [34,49-51]. Furthermore, the use of augmented feedback may provide extrinsic information during training sessions and plays a crucial role in enhancing motivation in individuals with stroke [18].

Usability feedback from 5 patients with chronic stroke indicated the multimodal exergame was easy to use, appropriately challenging, and included exercise movements that were suitable and easy to follow. The results were in line with the previous study [34,52] and suggested that the positive feedback from users unfamiliar with new technologies may be due to the user-friendly interface and appropriately challenging task levels incorporated into the intervention design. However, some users reported occasional delays in game graphics and sensor detection; exercise movements, such as alternating between toe taps and heel taps, were too complicated, and some instructions were unclear. In response to negative feedback from participants, technical adjustments were implemented to enhance sensor performance and reduce system latency. Additionally, minor revisions to graphics and certain instructions were made based on consensus from the expert panel and research team. The finalized version of the intervention program was developed through a rigorous process. Consequently, the preliminary prototype of the multimodal exergame demonstrated satisfactory scientific validity, aligning with constructs relevant to specific clinical and research objectives in poststroke rehabilitation.

The American Stroke Association demonstrated that encouraging physical activity among survivors of stroke should focus on moderate-intensity of exercises and managing risk factors to prevent secondary stroke and improve cardiovascular health [6]. To confirm exercise intensity, this study used the Borg scale for subjective assessment [53]. It has been documented that the RPE is a simple and practical method that does not require special equipment to measure an individual’s subjective level of exertion and is used to ensure that exercise intensity remains within tolerable limits [6,53]. The results of this study showed that the average perceived exertion rating during exercise was 13.40, indicating moderate-intensity exercise. Previous studies showed that exercise at moderate intensity promoted health benefits in patients with stroke [54,55]. Thus, the prototype of multicomponent exercise at moderate intensity could provide opportunities to support stroke recovery. Additionally, integrating objective measures (eg, heart rate monitoring via wearable devices) would further enhance the precision of intensity monitoring and enable more personalized and adaptive progression of the exercise program. Furthermore, combining perceived exertion with heart rate–based targets may facilitate automated adjustment of exercise intensity across sessions.

Emerging evidence suggests that multimodal interventions integrating physical and cognitive components may confer broader benefits than single-domain training in older adults [56,57]. In addition, research evidence has demonstrated that multimodal exercise improved health outcomes in older adult patients with stroke and frailty [7] and older adults with dementia [8]. In this feasibility study, participants demonstrated improvements in global cognition and executive function following a 6-week home-based multimodal exergame. However, these findings should be interpreted with caution due to the small sample size and the absence of a control group, which limit causal inference and increase the risk of overestimating intervention effects.

The observed trends in cognitive outcomes may be partly explained by the design of the intervention. Specifically, the exergame tasks were structured to engage multiple domains of cognition, with a particular emphasis on executive processes such as attention, inhibition, and cognitive flexibility. In contrast, memory-related tasks were less frequently represented and primarily limited to working memory components. In addition, the progression from lower to higher difficulty levels incorporated increased cognitive load, shorter response times, and greater exposure to distracting stimuli. These features likely required maintained attention and adaptive control, which may preferentially stimulate executive function [58,59]. Nevertheless, given the exploratory nature of this study, these mechanistic interpretations remain speculative and should be confirmed in controlled trials.

In contrast to the cognitive findings, motor performance, as assessed by the SPPB and TUG, did not show significant changes following the intervention. This lack of improvement may be related to several characteristics of the training program. To prioritize the safety of home-based and remote training, the intervention primarily emphasized seated activities and sit-to-stand exercises, with limited incorporation of dynamic balance, gait training, or higher-intensity lower extremity tasks. As a result, the stimulus may not have been sufficiently specific or intense to elicit measurable improvements in functional mobility outcomes that require integrated components of balance, walking speed, and transitional movements. Furthermore, the outcome measures themselves may have influenced the findings. Both SPPB and TUG capture complex, whole-body functional performance, including gait and dynamic balance [41-43], which were not extensively targeted in the intervention. Therefore, the transfer of training effects to these measures may have been limited. It is also possible that the relatively short intervention duration and limited sample size reduced the statistical power to detect changes in both motor performance and memory outcomes. Importantly, no adverse events were reported during training with the home-based multimodal exergame. After evaluating the feasibility of the program, the total PACES score was maintained over the training period, and a high adherence rate was presented. This finding highlights the safety and feasibility of remote, technology-assisted interventions for individuals with chronic stroke. Additionally, participants were able to engage consistently with the exergame protocol in remote areas and their home environments.

Taken together, while the findings support the feasibility of a home-based multimodal exergame intervention and suggest potential cognitive benefits, they should be interpreted as preliminary. Future studies using randomized controlled designs, larger sample sizes, and longer intervention periods are needed to establish efficacy. Additionally, incorporating more task-specific and higher-intensity motor components, as well as selecting outcome measures that are closely aligned with the training stimuli, may help clarify the extent to which such interventions can improve both cognitive and motor function in individuals with chronic stroke.

This study has several limitations that should be acknowledged. All participants demonstrated normal range of cognitive function, as determined by a screening tool. In addition, most participants had an educational background of high-school level or higher. Therefore, the findings primarily reflect the experiences of this specific group and may not capture the perspectives of individuals with different cognitive profiles or those with lower educational attainment. Although RPE is well-documented as a practical method for measuring an individual’s subjective level of exertion during exercise, incorporating objective measurements (eg, polar heart rate monitors) would provide a more precise and comprehensive assessment of exercise intensity, while enabling personalized intensity adjustment in future studies. Furthermore, participants’ technology literacy was not formally assessed and may have influenced the study outcomes, particularly usability, satisfaction, and adherence. As this study primarily assessed the content validity, usability, and feasibility of the intervention program in patients with chronic stroke, the effectiveness of the system and program requires further investigation through large-scale, methodologically rigorous clinical trials to establish definitive clinical outcomes.

The home-based multimodal exergame demonstrated acceptable content validity, usability, and feasibility in individuals with stroke. Unlike conventional home exercise programs, this multimodal exergame integrates both physical and cognitive training into an interactive home-based and remote rehabilitation platform for patients with chronic stroke. The intervention was found to be enjoyable, accessible, user-friendly, and safe, while showing potential benefits for cognitive outcomes, maintaining exercise enjoyment, and promoting exercise adherence throughout the training period. This preliminary evaluation suggests that the home-based multimodal exergame is a promising rehabilitation tool that may enhance patient engagement, support long-term rehabilitation participation, and help address barriers to stroke rehabilitation, particularly for individuals living in underserved or remote areas.