Balance control impairments are common in people with multiple sclerosis (MS) [1] and may affect approximately 75% of people with MS during the course of the disease [1,2]. Weakness, spasticity, and fatigue are just some of the factors that cause balance disorders; moreover, central sensorimotor integration of visual, proprioceptive, and vestibular inputs is a critical aspect of balance control that may underlie dysfunction in static and dynamic balance [3,4]. Imbalance is usually characterized by the reduced ability to maintain position and to move toward the limits of stability and delayed responses to postural adjustments [1,4]. These abnormalities, together with other impairments and disabilities, often prevent people with MS from correctly and easily performing daily living activities and are considered as a primary fall risk in MS [5] with many adverse physical and psychological consequences (eg, injuries, fear of falling, and loss of autonomy) [6,7].

Mainly for these reasons, MS is associated with physical inactivity [8,9] that may lead to rapid symptom worsening and increased risk of secondary conditions (eg, fractures, high rates of depression, and fatigue). In turn, these may seriously affect participation in societal interactions and ultimately translate into reduced quality of life (QoL) [10]. Thus, balance exercises have increasingly been recommended in MS to control symptoms, enhance participation in daily activities, and improve QoL [11]. Balance exercises can be considered as the cornerstone of MS interventions [4].

Pilates is a popular alternative method for the maintenance and improvement of balance [12]. It is a precisely controlled form of exercise that uses stabilizing muscles of the body. Specifically, Pilates uses a holistic approach: concentration, control, centering, flowing movement, precision, and breathing are 6 fundamental principles; if correctly followed, they can improve flexibility, strength, core stability, muscle control, breathing, and posture and increase body awareness with less ground impact and joint stress [12,13]. These aspects and the possibility to perform exercises at the most convenient intensity levels (ie, difficulty adjusted to own level of conditioning) make Pilates an ideal approach to treat physical symptoms of MS [14]. More specifically, although high-quality studies are needed owing to the methodological flaw of most of the current literature, a recent meta-analysis showed that Pilates exercises might be an optional effective method to treat balance in people with MS [15].

In this context, the advancement of technology makes new alternatives available to deliver rehabilitative interventions in the form of internet- and game-based exercises (exergames) [16]. Playing exergames is a form of whole-body physical exercise delivered via commercial nonimmersive virtual reality [17] that requires users to complete the assigned tasks aimed at improving fitness and promoting an active lifestyle [18]. Exergames stimulate users to be more active, allowing repetitive practice and real-time interaction with feedback and tracking of progress, which enhance motivation, treatment adherence, and training outcomes [19-21]. Games that are specifically designed to address a problem or teach a certain ability have had a lot of success [22]. Researchers have demonstrated that exergames are comparable with traditional therapies and can be more fun and acceptable [23].

In the field of neurological diseases, a recent meta-analysis [24] explored the extent to which exergaming was superior to physical therapy or other forms of rehabilitation in improving balance dysfunction. Results were consistent with those of previously published meta-analyses and systematic reviews supporting the efficacy of exergames in the neurological setting [25-28]. A beneficial effect of exergames was found in stroke; Parkinson disease; and to a lesser extent, in MS, whereas conclusions on other conditions, such as mild cognitive impairment or Alzheimer disease, traumatic brain injuries, and myelopathies, were not possible. Furthermore, the effect size of exergames on balance significantly depended on the weekly frequency of sessions, single session duration, and overall intervention duration, whereas it was independent of the setting (supervised or home-based) and strategy (exclusive or add-on) of the intervention. Finally, exergames were found to be safe, with only mild to moderate musculoskeletal adverse events found in a minority of patients, with no significantly increased risk of falls [29].

Among the internet- and game-based technologies used for exercise training, the Kinect motion sensor (Microsoft Corp) constitutes a high-potential approach; it provides a precise markerless, full-body 3D motion tracker with real-time feedback about user performance and enables users to virtually interact hands-free with a computer system without needing a remote controller [30-35]. Lozano-Quilis et al [36] demonstrated the effectiveness of RemoviEM, a Kinect-based system offering people with MS an intuitive and motivating way to perform weight transfer and upper limb coordination exercises. Results showed great improvement in balance outcomes in the group using RemoviEM compared with the control group receiving the same types of exercises in a standard rehabilitative intervention. More recently, Molhemi et al [37] assessed the efficacy of a rehabilitative protocol including standing, walking, and weight shifting exercises delivered through a Kinect-based versus conventional training on the improvement of balance and mobility in people with MS. Both interventions improved balance and mobility; however, training with exergames was more efficacious in enhancing cognitive-motor function and reducing falls, whereas conventional exercises led to better directional control.

In the next years, new exergames will become available to the therapists, and it is hoped that these will allow for the integration in MS outpatient and at-home settings to optimize patient clinical outcomes and to ensure the continuity of care [38]. On the basis of the usability challenges faced by people with MS, programmers should develop user-friendly interfaces customized for people with MS, which is the main assumption for high acceptance, adherence, and effectiveness.

The aim of this study was 2-fold: first, to design and develop a usable Kinect-based exergame implementing Pilates exercises appropriately adapted to MS requirements; second, to evaluate the usability in people with MS for future research and clinical use. An Italian network of experts in the field of MS rehabilitation was involved in the different phases of the study. The first steps were dedicated to literature revision and exercise identification, followed by prototyping based on co-design with MS health care professionals and people with MS. Usability testing was conducted on the final version of the tool.

This study consists of 2 parts: the design and development of an exergame implementing Pilates exercises and its customization for the needs of people with MS (part 1) and evaluation of its usability in terms of ease of use, playability, enjoyment, satisfaction, and acceptance (part 2).

The exergame was developed for asynchronous telerehabilitation purposes. The patient should be clinically supervised by a physical therapist for the evaluation of the progress of training and performances (Table 1). Moreover, for patients with neurological problems, no limitations in terms of session frequency and duration are considered; the more use, the higher the effect size of the exergame [24].

The study was approved by the regional ethics committee of the principal investigator (Azienda Ospedaliera San Martino Genoa; n° 134/2018). All participants gave written informed consent before entry into the study in accordance with the revised Declaration of Helsinki [39].

Part 2 was deputed to evaluate the usability of MS-FIT (release 1.0) with people with MS; an ad hoc questionnaire was used to analyze the ease of use, playability, enjoyment, satisfaction, and acceptance.

MS-FIT is a tool for individual, internet-based, game-based training using Microsoft Kinect Sensor V2. The high number of implemented exercises (Multimedia Appendix 1) and the different levels of difficulty prevent boredom.

The MS-FIT context is travel around the world to visit some of the most important cities (Figure 2 and Multimedia Appendix 4) and to learn the aspects of their culture through pictures (eg, landscape) and stories (eg, curiosity; Textbox 1). At each stay of the travel, a teacher of Pilates, represented by an avatar, shows and explains to the user the exercise they will have to perform; a related caption is displayed. The teacher’s avatar (a human model wearing web-based garments) is represented on the left of the screen, whereas the user’s avatar (a human model in a density gradient visualization) is represented on the right and allows real-time feedback to the participant (Figure 3).

Accessibility was guaranteed through the implementation of exercises tailored for people with MS and owing to the use of auditory instructions provided by the teacher’s avatar before starting the exercise while showing the related captions.

The challenge for the user is the execution of the exercises as correctly and accurately as possible at the same time as the teacher; green arrows (or circles for circular gestures), applied to the user’s body joint and segment mainly involved in the exercise, are displayed to suggest the correct direction and to guide the execution (Figure 3). For each correct gesture, the user is rewarded with web-based coins visible while gaming; the reward represents the feedback about performances given to the participant while playing MS-FIT. The accumulated web-based coins can be used to unlock the next stay (ie, next city) and progress along the travel; this progress represents another feedback to the user. Besides this modality of game progress, MS-FIT also provides several missions; for example, twice-daily trainings for 5 consecutive days allow completing the mission, Methodic (Multimedia Appendix 5).

As the game progresses, the user develops more skills that motivate them to continue the training; according to the travel progress, the level of difficulty in the next city increases in terms of the number of consecutive repetitions of the same exercise or in terms of exercise complexity; however, the user can also repeat the exercises of previously unlocked cities without training at increasing levels of difficulty.

All these aspects of MS-FIT are expected to be crucial for the engagement of the users and the success of the product.

MS-FIT collects use data and exercise metrics about the frequency and duration of exercise sessions, types of exercises that are performed or skipped by the user, performance in each single exercise, and other statistics (eg, the ratio of unlocked cities to total cities and the ratio of completed missions to total missions). However, these in-game metrics are not shown to the user and were not investigated in this usability study. All data are anonymized and sent to a central server for storage.

In this study, we aimed to design, develop, and user test MS-FIT, an internet-based, game-based tool deputed to deliver Pilates exercises for people with MS. Study outcomes support the therapeutic use of Pilates in MS management because it is a safe, active treatment method (few adverse effects) with high adherence (low dropout rate) and can improve important parameters in the target population, such as balance, gait, physical functional capacities, and even cognitive functions [69].

MS-FIT was formulated by integrating the principal knowledge and existing evidence from MS literature and by subjecting the exergame to critical appraisal from patients and experts. We adopted a research-based and iterative development process that considered the knowledge gained from different disciplines (human movement science, rehabilitation, physical activity, game design, and research) and users’ perspectives (people with MS and therapists) to holistically generate a potentially attractive and effective user-centered exergame for people with MS.

Although the participants had few experiences with exergames/video games and consoles, they became autonomous in using MS-FIT after only 3 sessions. All of them (9/9, 100%) completed the three 30-minute sessions and reported finding the tool easy to use, playable, enjoyable, and satisfactory. This result was corroborated by the quantity of coins gained and the number of cities unlocked, which could be considered as a surrogate index of the ease to use, playability, and acceptance. Feedback and suggestions provided by people with MS and therapists allowed for increasing the number of implemented exercises and, overall, improving the quality of the MS-FIT exergame.

In MS-FIT, we used Microsoft Kinect because it has been reported to be well accepted by users and for its effectiveness in several rehabilitative interventions for people with MS. For example, RemoviEM [36] is an intuitive and motivating way to train weight transfer and upper limb coordination of people with MS and to improve balance and mobility. Similarly, Molhemi et al [37] demonstrated the efficacy of a Kinect-based rehabilitative tool to train standing, walking, and weight shifting in MS. In addition, more consolidated experiences with a portable, low-cost tool such as the Nintendo Wii Balance Board could pave the way for an easy future adoption of Kinect-based exergames into clinical practice [24]. In this context, Brichetto et al [70] investigated the effectiveness of balance training based on the Nintendo Wii Balance Board versus a traditional rehabilitative training. Despite the excellent compliance with both interventions, a high clinically significant improvement in balance was shown in the group using the exergame. In addition, Prosperini et al [71] showed that training with the Nintendo Wii Balance Board induced microstructure modifications of the superior cerebellar peduncles, suggesting that high-intensity and task-oriented exercises such as those implemented in the Nintendo Wii Fit Plus package could favor brain reorganization in people with MS. Overall, internet-based, game-based training is an effective and efficient approach for empowering user engagement, which contributes to positive outcomes of an intervention [72,73].

To limit high dropout rates owing to potential low engagement with modern digital technology, the UCD approach, which incorporates game design principles (ie, goals, rules, feedback, points, time, reward structures, levels, and esthetics), was used in the process of designing and developing MS-FIT. Many studies demonstrated the benefits of using the UCD concept for developing exergames. For example, Lange et al [18] used the UCD approach to develop an internet-based, game-based program aimed at training the dynamic balance of individuals who had experienced a stroke. Hemingway et al [43] used the UCD process to develop a mobile game to influence the behavior of HIV service uptake among a key population. Howes et al [74] also used UCD to develop a bespoke exergame to deliver strength and balance exercise programs to older adults. Similarly, Kamnardsiri et al [34] developed and tested an internet-based, physical-cognitive, game-based training system using Kinect for older individuals. Therefore, the UCD approach is a core process that should be embedded in health game development to ensure usability and acceptability of and full benefit from exergame technology.

The findings from this study seem to be confirming these results. Consistent with previous studies, our game system provides feedback, including scores and performance outcomes, which enhances the motivation of the users [75,76]. However, it was not possible to evaluate the clinical benefit of MS-FIT because the number of user test sessions was very low; future planned studies will assess this aspect.

To the best of our knowledge, MS-FIT is the first exergame designed and developed to deliver Pilates exercises through an internet-based, game-based technology. Although MS-FIT was customized for people with MS, few adaptations would be necessary for use in the context of other clinical conditions [77-80].

The design, development, and usability testing of MS-FIT is the first step of a large research plan including an ongoing feasibility study followed by an RCT. The feasibility study has been planned to assess the satisfaction and acceptance toward the tool, adherence to the intervention, and dropout rate on a large sample size and for a long period of more intense use (ie, 5 sessions a week for 6 weeks). The use of validated questionnaires investigating usability [81-83] and the identification of potential bugs will be useful for further tool refinement. Moreover, the administration of clinical tests evaluating balance (eg, timed up-and-go test) will allow for estimating the effect of the intervention and its variance, which are necessary to calculate the appropriate sample size for the RCT. Finally, the feasibility study will allow the identification of groups of exercises with a level of difficulty that is suitable for the specific needs and capabilities of people with MS in different clinical stages.

An RCT (ClinicalTrials.gov; ID NCT04011579) will assess the effect of MS-FIT on physical and cognitive domains, mood, and QoL of people with MS. If RCT results are positive, MS-FIT could be proposed in combination with rehabilitation to potentiate the rehabilitative intervention effect or guarantee the continuity of care of people with MS.

Owing to the discontinuation of the Kinect Sensor V2, for the continued development of MS-FIT and successful translation into clinical practice, a new compatible commercial sensor must be identified and the app should be adapted. The Microsoft AzureKinect demonstrated sufficiently high interrater and intrarater reliability when compared with the Kinect Sensor V2. Moreover, a full-body skeletal tracking library in Python has recently become available. These aspects suggest that the Microsoft AzureKinect may be an adequate replacement within the MS-FIT system [84]. Another alternative solution to the Kinect Sensor V2 is MentorAge, a depth RGB (Red, Green, Blue) image sensor tracking system that operates on the Android system and uses infrared 3D capturing technology. It can detect a maximum of 4 people in a single room within a range of 0.6 to 5 m and has proven its capabilities and potentialities in real-life scenarios [80,85,86]. No additional equipment is thought to be used compared with the current version of MS-FIT.

Some limitations of the study need to be acknowledged. First, the evaluation of the target user’s experience involved a small number of participants. Furthermore, the usability results were obtained from a few training sessions. Thus, the findings should be considered as preliminary and interpreted with caution. Second, the minimum requested distance between the Kinect sensor and user could limit at-home use. New versions of the Kinect sensors could allow overcoming this limitation. Finally, this study investigated the usability by using an ad hoc questionnaire; however, in a future feasibility study, the use of validated questionnaires should be encouraged.

MS-FIT is an exergame using the Microsoft Kinect sensor for outpatient and at-home MS balance interventions. It digitally implements Pilates exercises for breathing, posture, and balance. MS-FIT is usable and well accepted by the target users, with positive feedback. Thus, the exergame is a promising tool for enhancing the motivation among people with MS to engage in exercise with the aim of improving their physical status.