Trusted Intelligent Sensing for Sport–Medicine Integrated Physical Activity Interventions: A Framework and Research Agenda
Year: 2026
Type: article
Abstract: Physical inactivity is among the largest modifiable contributors to the global burden of cardiovascular disease, type 2 diabetes, and several cancers, yet one in three adults worldwide still does not meet minimum activity recommendations (Guthold et al., 2024). Sports promotion and medical care have historically operated as separate systems: the sports sector delivers activity opportunities without clinical follow-up, while the medical sector treats the consequences of inactivity without prescribing exercise as a therapy. “Sport–medicine integration” — the deliberate coupling of physical-activity promotion with clinical and public-health pathways — is now a stated policy goal, but its delivery has been limited by the lack of trustworthy measurement and feedback between the two sectors. This Perspective argues that high-assurance intelligent sensing — wearable inertial and photoplethysmographic sensors, on-device activity recognition, and privacy-preserving data pipelines — is the missing technical layer that can make sport–medicine integration operable. We define six requirements for “high assurance”: clinical-grade accuracy across populations, transparent and human-auditable algorithms, privacy-preserving data handling, fairness across demographic groups, regulatory validity, and a closed-loop feedback path back to both the user and the clinician. We propose a four-layer framework — sensing, assurance, prescription, and feedback — and map the current evidence against it. The field’s priority is not more activity trackers but prospective, population-validated intervention trials in which sensing data drive tailored exercise prescriptions and are evaluated against health outcomes rather than step counts alone.
DOI: 10.64904/20261006
Authors: Chad Jensen, Anthony Murphy, Liam Morley
Institution: University of Alberta
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7225629447