The 2nd Biomechanics in Sport and Ageing Symposium: Artificial Intelligence (AI)’ is approaching rapidly. The symposium will take place at the Hungarian University of Sports Science, Budapest, Hungary, on 15-16 October 2024.
It is just timely to provide a perspective on the role of AI in health sciences, including sport and aging.
AI relies on computers to execute commands that historically required human intelligence. Now, we have apps that offer talking digital assistant services, respond to voice and text requests, it can answer questions, write poetry, generate images, draft emails, analyze personal photos, set a timer or place a phone call. Indeed, AI builds computational units that mimic human intelligence and abilities: language, communication, comprehension of concepts, automated thinking, (machine) learning, computer vision, and movements via robots [1]. Supervised or unsupervised machine learning, i.e., the science of coding computers to learn and behave as humans do, as a subset of AI, allows us to discern data patterns and structure. Deep learning optimizes supervised learning and trains models to learn how to map an input to an expected output [2,3] (Fig. 1).
Fig. 1. Overview of AI algorithms.
Any new research method can be mysterious. It is not easy to identify ways the new method can be used. In sport science, the number of publications key-worded ‘artificial intelligence’ has increased from 30 in 2014 to 400 in 2024. Some examples to highlight the proliferation of AI in sport science to demystify its use: to determine the nationality of the fastest 100-mile ultra-marathoners; identification of foot strike pattern using accelerometry towards injury prevention; analyzing weightlifting technique in novices; assessing the risk of severe mental distress among college students with respect to demographics, eating habits, lifestyles, and sport habits; predicting failure of anterior cruciate ligament after surgery in athletes returning to competition; predicting ground reaction forces during running from wearable ultrasound sensors; using dual-task tandem walking in the identification of sport-related concussions; predicting football injuries; training load identification during football; morphometrics to predict 20-m sprint performance in children.... Specifically, the Kitman Labs, a data firm works with several professional sports leagues with offices in Silicon Valley and Dublin.
They aim to gather milliard on-field data point and mine these vast data to predict injury for each individual athlete. Rocky Collis established the Mustard app. After submitting videos, computer vision, a form of AI, analyzes the images and recommends exercise routines. Zone7, with investors like New York Knicks forward Kristaps Porzingis, relies on AI-based data-analytics and managed to reduce injuries by a third over one season in players of Liverpool Football Club in the U.K. Zone 7’s AI model also compares a player’s output against data collected from other players: it has stored 5 million hours of athletes’ performance. Using AI, researchers look for patterns that led to injury.`4D Motion Sports in New Jersey, USA has been supporting the US figure skaters to reduce fatigue and improve technique. Paris-based SkillCorner, a sports-data firm, collects broadcast TV video from soccer leagues around the world for its 65 clients and runs the footages through an algorithm that tracks individual players’ location and speed. The data are used to derive performance metrics that aid player scouting and recruitment.
Within health sciences, gerontology research advocates the use of AI for diagnosis and treatment. Fatigue is a ubiquitous problem in aging and there is a huge effort to detect automatically fatigue using AI from gait in older adults with and without clinical conditions. Falls is a substantial healthcare cost item. Several labs are working on deep learning frameworks to detect an upcoming fall. Dementia cases increase worldwide and early detection, even as early as 10 years before onset, is a key priority for AI research in gerontology. We have collaborated extensively with various laboratories in using AI to classify neurological patients, older vs. younger adults, and fallers vs. non-fallers from acceleration data recorded by wearable devices while walking [4-6]. AI is used also in interventions to classify individuals who are more vs. less skilled in various sport activities, for example, in Tai Chi so that classes can be formed based on experience quantified objectively.
Outstanding speakers with numerous publications will address many of these issues at the 2nd Biomechanics in Sport and Ageing Symposium: Artificial Intelligence (AI)’. The lead keynote will introduce AI and the following two keynotes will provide state-of-the-art overviews of AI in sport and ageing. From 9 countries, 14 invited area-expert speakers will give (big) data-based examples for how AI is used in sport and ageing with respect to: Body structure and exercise prescription; Motor-cognitive function; Injury and disease, and Performance assessment and prediction in a workshop format. Register, learn, and enjoy the remarkable event!
References
- Turing, A.M. On computable numbers, with an application to the Entscheidungsproblem. Proc London Math Soc. 1936, 58, 230–265.
- Matsuo, Y.; LeCun, Y.; Sahani, M.; Precup, D.; Silver, D.; Sugiyama, M.; Uchibe, E.; Morimoto, J. Deep learning, reinforcement learning, and world models. Neural Netw. 2022, 152, 267-275.
- Zhang, A.; Wu, Z.; Wu, E.; Wu, M.; Snyder, M.P.; Zou, J.; Wu, J.C. Leveraging physiology and artificial intelligence to deliver advancements in health care. Physiol Rev. 2023, 103, 2423-2450.
- Rehman, R.Z.U.; Zhou, Y.; Del Din, S.; Alcock, L.; Hansen, C.; Guan, Y.; Hortobágyi, T.; Maetzler, W.; Rochester, L.; Lamoth, C.J.C. Gait Analysis with Wearables Can Accurately Classify Fallers from Non-Fallers: A Step toward Better Management of Neurological Disorders. Sensors (Basel). 2020, 20.
- Zhou, Y.; Romijnders, R.; Hansen, C.; Campen, J.V.; Maetzler, W.; Hortobágyi, T.; Lamoth, C.J.C. The detection of age groups by dynamic gait outcomes using machine learning approaches. Sci Rep. 2020, 10, 4426.
- Zhou, Y.; Zia Ur Rehman, R.; Hansen, C.; Maetzler, W.; Del Din, S.; Rochester, L.; Hortobágyi, T.; Lamoth, C.J.C. Classification of Neurological Patients to Identify Fallers Based on Spatial-Temporal Gait Characteristics Measured by a Wearable Device. Sensors (Basel). 2020, 20.
Written by Professor Tibor Hortobágyi
