Background: Childhood cancer survivors are at increased risk of cardiometabolic complications that are exacerbated by poor health behaviors. Critically, many survivors do not meet physical activity guidelines. Objective: The primary aim was to evaluate the feasibility and acceptability of iBounce, a digital health intervention for educating and engaging survivors in physical activity. Our secondary aims were to assess the change in survivors’ physical activity levels and behaviors, aerobic fitness, and health-related quality of life (HRQoL) after participating in the iBounce program. Methods: We recruited survivors aged 8 to 13 years who were ≥12 months post cancer treatment completion. The app-based program involved 10 educational modules, goal setting, and home-based physical activities monitored using an activity tracker. We assessed objective physical activity levels and behaviors using cluster analysis, aerobic fitness, and HRQoL at baseline and after the intervention (week 12). Parents were trained to reassess aerobic fitness at home at follow-up (week 24). Results: In total, 30 participants opted in, of whom 27 (90%) completed baseline assessments, and 23 (77%) commenced iBounce. Our opt-in rate was 59% (30/51), and most (19/23, 83%) of the survivors completed the intervention. More than half (13/23, 57%) of the survivors completed all 10 modules (median 10, IQR 4-10). We achieved a high retention rate (19/27, 70%) and activity tracker compliance (15/19, 79%), and there were no intervention-related adverse events. Survivors reported high satisfaction with iBounce (median enjoyment score 75%; ease-of-use score 86%), but lower satisfaction with the activity tracker (median enjoyment score 60%). Parents reported the program activities to be acceptable (median score 70%), and their overall satisfaction was 60%, potentially because of technological difficulties that resulted in the program becoming disjointed. We did not observe any significant changes in physical activity levels or HRQoL at week 12. Our subgroup analysis for changes in physical activity behaviors in participants (n=11) revealed five cluster groups: most active, active, moderately active, occasionally active, and least active. Of these 11 survivors, 3 (27%) moved to a more active cluster group, highlighting their engagement in more frequent and sustained bouts of moderate-to-vigorous physical activity; 6 (56%) stayed in the same cluster; and 2 (18%) moved to a less active cluster. The survivors’ mean aerobic fitness percentiles increased after completing iBounce (change +17, 95% CI 1.7-32.1; P=.03) but not at follow-up (P=.39). Conclusions: We demonstrated iBounce to be feasible for delivery and acceptable among survivors, despite some technical difficulties. The distance-delivered format provides an opportunity to engage survivors in physical activity at home and may address barriers to care, particularly for regional or remote families. We will use these pilot findings to evaluate an updated version of iBounce. Trial Registration: Australian New Zealand Clinical Trials Registry ACTRN12621000259842; https://anzctr.org.au/Trial/Registration/TrialReview.aspx?ACTRN=ACTRN12621000259842

To evaluate the impact of Health interventions promoting physical activity, researchers typically conduct pre- and post-assessments using accelerometers. While aggregated metrics such as daily counts, daily steps and time spent at various intensity levels are commonly used, very few studies exploit the richness of the data often collected with a very fine granularity. We investigate the benefit of a deeper analysis of wrist accelerometry data to understand physical activity behaviours throughout the day, as well as how these may change overtime. To analyse physical activity behaviour changes, we propose a methodology that extracts bouts of physical activity characterised by their activity levels and duration, and uses these as features to cluster participants’ daily and hourly behaviours. We then compare these clusters to assess changes following an intervention promoting physical activity in children. We demonstrate that this approach provides a more insightful analysis of the physical activity behaviours because it highlights the nature and the timing of behaviour changes, when present. We illustrate this methodology using data from research-grade activity trackers (GENEActiv) and explain the insights discovered in the context of an intervention aimed at educating school children about healthy behaviours.

Background: Sensors are increasingly used in health interventions to unobtrusively and continuously capture participants’ physical activity in free-living conditions. The rich granularity of sensor data offers great potential for analyzing patterns and changes in physical activity behaviors. The use of specialized machine learning and data mining techniques to detect, extract, and analyze these patterns has increased, helping to better understand how participants’ physical activity evolves. Objective: The aim of this systematic review was to identify and present the various data mining techniques employed to analyze changes in physical activity behaviors from sensors-derived data in health education and health promotion intervention studies. We addressed two main research questions: (1) What are the current techniques used for mining physical activity sensor data to detect behavior changes in health education or health promotion contexts? (2) What are the challenges and opportunities in mining physical activity sensor data for detecting physical activity behavior changes? Methods: The systematic review was performed in May 2021 using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. We queried the Association for Computing Machinery (ACM), IEEE Xplore, ProQuest, Scopus, Web of Science, Education Resources Information Center (ERIC), and Springer literature databases for peer-reviewed references related to wearable machine learning to detect physical activity changes in health education. A total of 4388 references were initially retrieved from the databases. After removing duplicates and screening titles and abstracts, 285 references were subjected to full-text review, resulting in 19 articles included for analysis. Results: All studies used accelerometers, sometimes in combination with another sensor (37%). Data were collected over a period ranging from 4 days to 1 year (median 10 weeks) from a cohort size ranging between 10 and 11615 (median 74). Data preprocessing was mainly carried out using proprietary software, generally resulting in step counts and time spent in physical activity aggregated predominantly at the daily or minute level. The main features used as input for the data mining models were descriptive statistics of the preprocessed data. The most common data mining methods were classifiers, clusters, and decision-making algorithms, and these focused on personalization (58%) and analysis of physical activity behaviors (42%). Conclusions: Mining sensor data offers great opportunities to analyze physical activity behavior changes, build models to better detect and interpret behavior changes, and allow for personalized feedback and support for participants, especially where larger sample sizes and longer recording times are available. Exploring different data aggregation levels can help detect subtle and sustained behavior changes. However, the literature suggests that there is still work remaining to improve the transparency, explicitness, and standardization of the data preprocessing and mining processes to establish best practices and make the detection methods easier to understand, scrutinize, and reproduce.

Wearable accelerometers record physical activity with high resolution, potentially capturing the rich details of behaviour changes and habits. Detecting these changes as they emerge is valuable information for any strategy that promotes physical activity and teaches healthy behaviours or habits. Indeed, this offers the opportunity to provide timely feedback and to tailor programmes to each participant’s needs, thus helping to promote the adherence to and the effectiveness of the intervention. This article presents and illustrates U-BEHAVED, an unsupervised algorithm that periodically scans step data streamed from activity trackers to detect physical activity behaviour changes to assess whether they may become habitual patterns. Using rolling time windows, current behaviours are compared with recent previous ones, identifying any significant change. If sustained over time, these new behaviours are classified as potentially new habits. We validated this detection algorithm using a physical activity tracker step dataset (N = 12,798) from 79 users. The algorithm detected 80% of behaviour changes of at least 400 steps within the same hour in users with low variability in physical activity, and of 1600 steps in those with high variability. Based on a threshold cadence of approximately 100 steps per minute for standard walking pace, this number of steps would suggest approximately 4 and 16 min of physical activity at moderate-to-vigorous intensity, respectively. The detection rate for new habits was 80% with a minimum threshold of 500 or 1600 steps within the same hour in users with low or high variability, respectively.

BACKGROUND Sensors are increasingly used in health interventions to unobtrusively and continuously capture participants’ physical activity in free-living conditions. The rich granularity of sensor data offers great potential for analyzing patterns and changes in physical activity behaviors. The use of specialized machine learning and data mining techniques to detect, extract, and analyze these patterns has increased, helping to better understand how participants’ physical activity evolves. OBJECTIVE The aim of this systematic review was to identify and present the various data mining techniques employed to analyze changes in physical activity behaviors from sensors-derived data in health education and health promotion intervention studies. We addressed two main research questions: (1) What are the current techniques used for mining physical activity sensor data to detect behavior changes in health education or health promotion contexts? (2) What are the challenges and opportunities in mining physical activity sensor data for detecting physical activity behavior changes? METHODS The systematic review was performed in May 2021 using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. We queried the Association for Computing Machinery (ACM), IEEE Xplore, ProQuest, Scopus, Web of Science, Education Resources Information Center (ERIC), and Springer literature databases for peer-reviewed references related to wearable machine learning to detect physical activity changes in health education. A total of 4388 references were initially retrieved from the databases. After removing duplicates and screening titles and abstracts, 285 references were subjected to full-text review, resulting in 19 articles included for analysis. RESULTS All studies used accelerometers, sometimes in combination with another sensor (37%). Data were collected over a period ranging from 4 days to 1 year (median 10 weeks) from a cohort size ranging between 10 and 11615 (median 74). Data preprocessing was mainly carried out using proprietary software, generally resulting in step counts and time spent in physical activity aggregated predominantly at the daily or minute level. The main features used as input for the data mining models were descriptive statistics of the preprocessed data. The most common data mining methods were classifiers, clusters, and decision-making algorithms, and these focused on personalization (58%) and analysis of physical activity behaviors (42%). CONCLUSIONS Mining sensor data offers great opportunities to analyze physical activity behavior changes, build models to better detect and interpret behavior changes, and allow for personalized feedback and support for participants, especially where larger sample sizes and longer recording times are available. Exploring different data aggregation levels can help detect subtle and sustained behavior changes. However, the literature suggests that there is still work remaining to improve the transparency, explicitness, and standardization of the data preprocessing and mining processes to establish best practices and make the detection methods easier to understand, scrutinize, and reproduce.