FİZİKSEL AKTİVİTEDE GİYİLEBİLİR YAPAY ZEKA YAKLAŞIMI: SİSTEMATİK LİTERATÜR TARAMASI

WEARABLE ARTIFICIAL INTELLIGENCE APPROACH IN PHYSICAL ACTIVITY: A SYSTEMATIC LITERATURE REVIEW

Abstract

The purpose of this study is to examine artificial intelligence (AI)-supported physical activity applications through wearable sensors. Specifically, it aims to evaluate the contribution of AI to the objective measurement of physical activity levels and the potential of these technologies to increase physical activity in sedentary individuals. Additionally, it seeks to systematically review the literature to understand how AI-based applications developed in recent years impact physical activity. To this end, a systematic literature review was conducted in the Scopus, Web of Science, and PubMed databases using the keywords “physical activity,” “artificial intelligence,” “wearable sensor,” and “wearable artificial intelligence.” The inclusion criteria for the study were: articles published between 2015 and 2024, written in English, with open access permissions, falling under the category of sports sciences, measuring physical activity using wearable sensors, or providing AI-supported feedback and examining its impact on physical activity. Additionally, studies that incorporated wearable AI and included at least one assessment measuring physical activity were also part of the inclusion criteria. Publications outside of these criteria and review-type studies were excluded. The search yielded a total of 582 articles, and after evaluations based on the inclusion and exclusion criteria, 17 articles were included in the study. It is anticipated that wearable sensors and AI-supported physical activity measurement will enable a more objective evaluation of physical activity programs, particularly for sedentary individuals. Therefore, it is emphasized that AI-supported physical activity designs should be further developed.

Keywords: Physical activity, artificial intelligence, wearable sensor, wearable artificial intelligence.

Öz

Bu çalışmanın amacı, giyilebilir sensörler aracılığıyla yapay zeka destekli fiziksel aktivite uygulamalarını incelemektir. Özellikle, yapay zekanın fiziksel aktivite düzeylerini objektif olarak ölçmedeki katkısını ve bu teknolojilerin sedanter bireylerde fiziksel aktiviteyi artırmadaki potansiyelini değerlendirmeyi hedeflemektedir. Ayrıca, son yıllarda geliştirilen yapay zeka tabanlı uygulamaların fiziksel aktiviteyi nasıl etkilediğini anlamak için literatürü sistematik olarak gözden geçirmeyi amaçlamaktadır. Bu doğrultuda, Scopus, Web of Science ve PubMed veri tabanında “physical activity,” “artificial intelligence,” “wearable sensor” ve “wearable artificial intelligence” anahtar kelimeleri kullanılarak sistematik bir literatür taraması gerçekleştirilmiştir. Çalışmaya dahil edilme kriterleri; 2015-2024 yılları arasında yayımlanmış, İngilizce dilinde, açık erişim izni bulunan, spor bilimleri kategorisine giren, giyilebilir sensörler kullanılarak fiziksel aktiviteyi ölçen veya yapay zeka destekli geri bildirim sağlayan araştırmalar ile fiziksel aktivite üzerindeki etkisini inceleyen çalışmalardır. Ayrıca, giyilebilir yapay zekayı içeren ve fiziksel aktiviteyi ölçen en az bir değerlendirme barındıran çalışmalar da dâhil edilme kriterleri arasında yer almaktadır. Bu kriterlerin dışında kalan ve derleme türünde olan yayınlar ise hariç bırakılmıştır. Tarama sonucunda toplam 582 makaleye ulaşılmış; dahil etme ve hariç bırakma kriterlerine göre yapılan değerlendirmeler sonucunda 17 makale çalışmaya dâhil edilmiştir. Giyilebilir sensörler ve yapay zeka destekli fiziksel aktivite ölçümünün, özellikle sedanter bireyler için fiziksel aktivite programlarının daha objektif değerlendirilmesini sağlayacağı öngörülmektedir. Bu nedenle, yapay zeka ile desteklenmiş fiziksel aktivite tasarımlarının geliştirilmesi gerektiği vurgulanmaktadır.

Anahtar Terimler: Physical activity, artificial intelligence, wearable sensor, wearable artificial intelligence.

KAYNAKÇA

Ahmed, A., Aziz, S., Qidwai, U., Farooq, F., Shan, J., Subramanian, M., Chouchane, L., EINatour, R., Abd-Alrazaq, A., Pandas, S., & Sheikh, J. (2023). Wearable artificial intelligence for assessing physical activity in high school children. Sustainability, 15(638). https://doi.org/10.3390/su15010638

Alam, R., Peden, D. B., & Lach, J. C. (2021). Wearable respiration monitoring: Interpretable inference with context and sensor biomarkers. IEEE Journal of Biomedical and Health Informatics, 25(6), 2292–2301. https://doi.org/10.1109/JBHI.2020.3035776

An, R., Shen, J., Wang, X., & Yang, Q. (2023). Artificial intelligence in physical activity interventions: Systematic review and future directions. Journal of Medical Internet Research, 25, e45108. https://doi.org/10.2196/45108

Beltrame, T., Amelard, R., Wong, A., & Hughson, R. L. (2018). Extracting aerobic system dynamics during unsupervised activities of daily living using wearable sensor machine learning models. Journal of Applied Physiology, 124(3), 473–481. https://doi.org/10.1152/japplphysiol.00299.2017

Capecci, M., Pournajaf, S., Galafate, D., Sale, P., Le Pera, D., Goffredo, M., De Pandis, M. F., Andrenelli, E., Pennacchioni, M., Ceravolo, M. G., & Franceschini, M. (2019). Clinical effects of robot-assisted gait training and treadmill training for Parkinson’s disease: A randomized controlled trial. Annals of Physical and Rehabilitation Medicine, 62, 303–312. https://doi.org/10.1016/j.rehab.2019.01.003

Chereshnev, R., & Kertész-Farkas, A. (2018). RapidHARe: A computationally inexpensive method for real-time human activity recognition from wearable sensors. arXiv. https://arxiv.org/abs/1809.09412

Galasso, S., Baptista, R., Molinara, M., Pizzocaro, S., Calabrò, R. S., & De Nunzio, A. M. (2023). Predicting physical activity levels from kinematic gait data using machine learning techniques. Engineering Applications of Artificial Intelligence, 123, 106487. https://doi.org/10.1016/j.engappai.2023.106487

Garcia-Tirado, J., Brown, S. A., Laichuthai, N., Colmegna, P., Koravi, C. L. K., Ozaslan, B., Corbett, J. P., Barnett, C. L., Pajewski, M., Oliveri, M. C., Myers, H., & Breton, M. D. (2021). Anticipation of historical exercise patterns by a novel artificial pancreas system reduces hypoglycemia during and after moderate-intensity physical activity in people with type 1 diabetes. Diabetes Technology & Therapeutics, 23(4), 271-280. https://doi.org/10.1089/dia.2020.0516

GTech. (2023). Yapay zeka nedir?. GTech Insights. https://gtech.com/yapay-zeka-nedir

Hassoon, A., Baig, Y., Naiman, D. Q., Celentano, D. D., Lansey, D., Stearns, V., Coresh, J., Schrack, J., Martin, S. S., Yeh, H.-C., Zeilberger, H., & Appel, L. J. (2021). Randomized trial of two artificial intelligence coaching interventions to increase physical activity in cancer survivors. npj Digital Medicine, 4, 168. https://doi.org/10.1038/s41746-021-00539-9

https://doi.org/10.2196/35867

Itoh, N., Mishima, H., Yoshida, Y., Yoshida, M., Oka, H., & Matsudaira, K. (2022). Evaluation of the effect of patient education and strengthening exercise therapy using a mobile messaging app on work productivity in Japanese patients with chronic low back pain: Open-label, randomized, parallel-group trial. JMIR mHealth and uHealth, 10(5), e35867.

Jacobs, P. G., Resalat, N., Hilts, W., Young, G. M., Leitschuh, J., Pinsonault, J., El Youssef, J., Branigan, D., Gabo, V., Eom, J., Ramsey, K., Dodier, R., Mosquera-Lopez, C., Wilson, L. M., & Castle, J. R. (2023). Integrating metabolic expenditure information from wearable fitness sensors into an AI-augmented automated insulin delivery system: A randomised clinical trial. The Lancet Digital Health, 5(8), e466–e474. https://doi.org/10.1016/S2589-7500(23)00112-7

Jeong, J.-H., Lee, I.-G., Kim, S.-K., Kam, T.-E., Lee, S.-W., & Lee, E. (2024). DeepHealthNet: Adolescent obesity prediction system based on a deep learning framework. IEEE Journal of Biomedical and Health Informatics, 28(4), 2282–2292.  https://doi.org/10.1109/JBHI.2024.3356580

Kobsar, D., & Ferber, R. (2018). Wearable sensor data to track subject-specific movement patterns related to clinical outcomes using a machine learning approach. Sensors, 18(9), 2828. https://doi.org/10.3390/s18092828

McGrath, R. L., Ziegler, M. L., Pires-Fernandes, M., Knarr, B. A., Higginson, J. S., & Sergi, F. (2019). The effect of stride length on lower extremity joint kinetics at various gait speeds. PLOS ONE, 14(2), e0211987. https://doi.org/10.1371/journal.pone.0211987

Mokmin, N. A. M. (2020). The effectiveness of a personalized virtual fitness trainer in teaching physical education by applying the artificial intelligent algorithm. International Journal of Human Movement and Sports Sciences, 8(5), 258–264. https://doi.org/10.13189/saj.2020.080514

Polechoński, J. (2024). Assessment of the intensity and attractiveness of physical exercise while playing table tennis in an immersive virtual environment depending on the game mode. BMC Sports Science, Medicine and Rehabilitation, 16, 155. https://doi.org/10.1186/s13102-024-00945-y

Tsai, C.-H., Christian, M., Kuo, Y.-Y., Lu, C. C., Lai, F., & Huang, W.-L. (2024). Sleep, physical activity and panic attacks: A two-year prospective cohort study using smartwatches, deep learning and an explainable artificial intelligence model. Sleep Medicine, 114, 55–63. https://doi.org/10.1016/j.sleep.2023.12.011

Wenzel, C., Liebig, T., Swoboda, A., Smolareck, R., Schlagheck, M. L., Walzik, D., Groll, A., Goulding, R. P., & Zimmer, P. (2024). Machine learning predicts peak oxygen uptake and peak power output for customizing cardiopulmonary exercise testing using non-exercise features. European Journal of Applied Physiology. https://doi.org/10.1007/s00421-024-05543-x

World Health Organization. (2024). Physical activity. https://www.who.int/news-room/fact-sheets/detail/physical-activity