Keywords

1 Introduction

According to the Centers for Disease Control and Prevention, ASD is one of the most commonly diagnosed disabilities for individuals in the United States and the fastest growing developmental disability that may result in significant social, communication, and behavioral challenges. ASD prevalence estimates increased from an average of 4 per 10,000 individuals in the mid-60s [1] to 1 per 59 in 2018Footnote 1. In addition, motor impairment is common in those with ASD [2, 3], who are also almost 1.5 times more likely to be overweight or obese compared to their typically developing peers [4]. These deficits may be exacerbated by their tendencies to adopt sedentary lifestyles and activities that lack social play [5]. A growing number of individuals with ASD may not possess the skills needed to be physically active and participate in sport and recreational activities [6]. While there is evidence indicating that physical activity has positive effects on the overall health (e.g., weight, motor skills) of individuals with ASD [7] as well as physical and cognitive benefits [8, 9], there is limited research in providing innovative environments to engage individuals with ASD in physical activity. This is further complicated by the difficulties experienced by individuals with ASD when estimating their relative efforts invested in physical activity. To this end, this research proposes an interactive mobile application: Heart Runner designed specifically for individuals with ASD to comprehend real-time information of their energy expenditures during physical activities by means of interactive heart rate visualization. Furthermore, the Heart Runner aims to provide an engaging environment where individuals with ASD can compete with one another in sustaining accelerated heart rates with the overall goal of stimulating extended durations of a physical activity. In the context of this paper, physical activity is defined as bodily movement that results in energy expenditure. We present preliminary evaluations of the Heart Runner with promising results suggesting increased heart rates for those who completed a given physical activity using the application compared to those who did not. This finding indicates the potential benefits of utilizing mobile applications in providing engaging environments and enhancing self-motivation for individuals with ASD in physical activity.

2 Related Work

A brief overview of related work is presented in this section, for detailed systematic reviews of the literature, see [8, 10]. Technology has been shown to have a positive effect on engagement of individuals with ASD when used simultaneously with exercise [11], suggesting predictable environments and visual stimulation created by technology may increase the potential for independent participation. As an established evidence-based practice (EBP), technology has also been shown to be highly motivating and reinforcing [12]. Prior experiments have demonstrated that through the use of technology-aided instruction, individuals with ASD are able to engage in more on-task behaviors and may learn physical activity skills at a faster rate than those without technology-aided instruction [13]. Notable applicationsFootnote 2 have demonstrated successful integrations of technology in enhanced physical activities though not always designed specifically for individuals with ASD. One observation is the notion of exergaming used in these applications, whereby gaming and technology are coupled to engage individuals in aerobic exercise, as body movements, reactions, and energy expenditure are tracked through participation in the game [14]. Exergaming is further investigated in [15], where findings have shown significant improvements in attention and working memory and decreases in stereotypical behaviors in individuals with ASD immediately after participating in a 20-min exergaming intervention. Motivated by these prior research findings, the Heart Runner implements the exergaming concept as discussed below.

3 Heart Runner: Design and Development

The Heart RunnerFootnote 3 is a cross-platform mobile application. The trial implementation shown in this paper runs on the Android operating system. It utilizes the NativeScriptFootnote 4 library to create interfaces and to access utility functions such as Bluetooth and phone cameras. The Heart Runner is compatible with heart rate monitors that support optical sensing technology when reading heart beats such as the Heart ZoneFootnote 5 heart rate monitor used in the trials presented in this paper. The Heart Runner bases its architecture on the Model-View-ViewModel design patternFootnote 6 that aims to support code reuse and extensibility. The application collects a user’s heart rate data in real-time during physical activity and visualizes changes to this person’s heart rate intensity as an avatar climbing up and down a mountain, as shown in Fig. 1a. The features shown in the Heart Runner follow several EBPs reported by the National Professional Development Center on ASDFootnote 7 that are specific to individuals with ASD, as discussed below.

Fig. 1.
figure 1

Heart Runner interface

Full size image

The mountain is segmented into four colored zones, where each zone indicates a particular range of heart rate intensity. More specifically, the red zone indicates light exercise (at least 50% but below 60% of recommended maximum heart rates), the yellow zone indicates moderate exercises (at least 60% but below 70% of recommended maximum heart rates), the green zone indicates vigorous exercise (at least 70% but below 90% of recommended maximum heart rates), and the white zone indicates a danger zone (90% or above recommended maximum heart rates). The recommended maximum heart rate is determined by the age of an individual, where for those 16 and over: Maximum Heart Rate = 220 − Current Age and for those under 16: Maximum Heart Rate = 208 − 0.7 × Current Age [16]. The Heart Runner then calculates an increase in one’s heart rates as a percentile value: Heart Rate Activity (%) = Current Heart Rate ÷ Maximum Heart Rate. The application allows a maximum of four players at a time with their respective avatars. As a player’s heart rates are elevated, this person’s avatar will synchronously climb up the mountain in the Hear Runner application in real-time. In addition, players can view their current heart rates in BPM (beats per minute) and collect stars as they spend a target amount of time in a particular zone on the mountain. For instance, a red star indicates a player has spent 30 s completing light exercise, a yellow start indicates a player has spent 15 s completing moderate exercise, and a green star indicates a player has spent 10 s completing vigorous exercise. Finally, Heart Runner provides a circular timer (top left corner of the screen) as a visual cue to assist players understanding the target duration of their exercise, and allows terminations of an exercise session at any given time (via the Stop option located at the top right corner of the screen). Once an exercise session has ended, a summary page presents an overview of a player’s heart rates recorded during the entire duration of the exercise. An example of such a summary page is shown in Fig. 1b. Summary pages are intended for caretakers to gain insights of the completed exercise, and to recommend or adjust desirable parameters accordingly such as a subsequent activity type and future target exercise durations.

4 Preliminary Evaluation

We conducted two trials of real-world user studies to evaluate whether the Heart Runner was effective in increasing heart rate intensity levels for individuals with ASD during physical activity. Trial one is a pilot study with the overall goal of testing technical feasibilities of the proposed application as well as refinement on experimental procedures. Trail two aims to overcome technical and procedure issues discovered in trail one and to provide additional results in the evaluation of the Heart Runner. Participants were drawn from different cohorts.

4.1 Trial One

A total of 20 individuals with ASD from a Southern California Transition school between the age of 18 and 21 took part in trial one. They were randomly assigned to join either the control group or the application group. Both groups were asked to exercise on stationary bicycles for a targeted duration at 20 min, with the only different being that those in the application group were asked to complete the exercise using the Heart Runner application. In the application group, the participants could select from several preset images as their avatars. Based on their age group discussed earlier, their recommended heart rate ranges would be 100–170 BPM with an optimal heart rate for vigorous activity reaching 144 BPM.

We experienced connection drops and range interferences for some participants during this trial, where their heart rate data had to be discarded as a result. The remaining dataset contains heart rates recorded for 15 participants, as shown in Fig. 2. We found that those in the application group (Fig. 2b) generated a higher average heart rate that is within the recommended ranges at 101.66 BPM, compared to those of the baseline group at approximately 90.63 BPM below the recommend ranges (Fig. 2a). The peak average heart rate in the baseline groups was at approximately 54.70%, which is outperformed by the application group at approximately 62.14%. Although none of the participants’ heart rates in trial 1 reached the optimal 144 BPM, we found the heart rates of the application group to be less dispersed compared to those of the baseline group, with standard deviations of 14.80 and 23.09 respectively. In addition, Fig. 2c presents a timeline view of the average heart rates generated for each group throughout the exercise in trial 1. The heart rate differences between the two groups are amplified throughout, where elevated heart rates are constantly shown for the application group, and particularly at the beginning such as the first 10 min of the exercise session.

Fig. 2.
figure 2

Trial 1 results

Full size image

To elicit further feedback from the participants regarding the usability of the Heart Runner application, participants were asked to complete the System Usability Scale (SUS) questionnaire [17]. Instead of Likert Scales, we used a 3-point Smiley Face Likert Scale [18] as recommended for children with medical conditions. The usability feedback from the participants was generally very positive where most cited the application was “easy to use”, “fun”, and therefore “would like to keep playing”. Although it may be necessary to recognize that not all participants were able to comprehend the purpose of the SUS questionnaire, since some commented on dislikes of sweating and the exercise itself. Nonetheless, the findings from trial 1 were encouraging and the experiment provided us an opportunity to identify areas of improvement.

4.2 Trial Two

We made changes to the usability questionnaire in trial 2, where participants were shown screenshots of four key features in the Heart Runner (the mountain, the avatars, the timer, and the star rewards) when providing feedback instead of statements for agreement/disagreement previously used in the SUS questionnaire. In addition, we provided upgrades to ensure successful connections before each session. Furthermore, we allowed participants to take photos of themselves as their avatars instead of choosing from a set of given images. A total of 12 individuals with ASD between the age of 18 and 21 took part in trial 2, where the same procedures used in trial 1 were applied.

Figure 3 presents the raw heart rate data collected from each participant in trail 2 for both groups. We found that those in the application group (Fig. 3b) generated a higher average heart rate that is within the recommended ranges at 107.11 BPM (StDev 16.12), compared to those of the baseline group at approximately 94.67 BPM (StDev 16.57) below the recommend ranges (Fig. 3a). The peak average heart rate in the baseline groups was at approximately 60.50%, which is outperformed by the application group at approximately 62.60%. Furthermore, Fig. 3c shows that the heart rate differences between the two groups are evident throughout the duration of the exercise, where elevated heart rates are constantly shown for the application group. Similar to what we found in trial 1, these differences appear at their largest at the beginning of the exercise session (e.g. approximately the first 8 min). We found that the avatar feature was most liked based on the usability feedback collected in trial 2. In particular, using a photo of oneself as the chosen avatar instead of using a given image was very well received by the participants, as they found it easier to locate themselves on the screen where there may be several other players in the same session. In addition, we received positive feedback on the mountain, the timer, and the star reward system on their helpfulness in conveying heart rate information in real-time.

Fig. 3.
figure 3

Trial 2 results

Full size image

5 Conclusions and Future Work

The evaluation results collected in both trials have provided preliminary evidence suggesting the potential benefits of visualizing heart rate data in real-time in an effort to potentially promote engagement and self-motivation among individuals with ASD during physical activity. In particular, we have consistently found higher average heart rates in those who used the Heart Runner application during physical activity compared to those who did not. Although these differences were not shown to be statistically significant (p>0.05) given the small sample sizes of both trials, further experiments may be necessary to verify this observation with larger number of participants. In addition, notable differences in heart rates of the two groups were especially found at the beginning of an exercise session, suggesting statistically significant differences may vary depending on a specific phase of physical activity. Future experiments may focus on measuring whether the Heart Runner application is effective at engaging individuals with ASD in extended exercise durations and at promoting user acceptance over a period of time. Moreover, the Heart Runner has provided a social environment and a reward system where players could compete with one another during a physical activity, where we have observed not only awareness but also excitement from the participants. Future studies may focus on evaluating the effectiveness of the social and reward element in generating motivation and engagement for individuals with ASD. Furthermore, the studies presented in this paper have been conducted using one type of physical activity, where future experiments involving other types of exercises may also be informative. Likewise, other themes in addition to mountains may be explored in future studies to identify other feasible and appropriate means to provide personalized visualizations of heart rate data to stimulate motivation and engagement from individuals with ASD during physical activity. Last but not least, user feedback collected from these trial experiments will inform development iterations of the Heart Runner to exercise user-centered design.