From data to diagnosis: A comprehensive review of machine learning-driven wearable sensors in healthcare.
Clinical Snapshot
PICO Framework
| P — Population | Human patients and individuals undergoing physiological monitoring across healthcare settings |
| I — Intervention | Machine learning (ML)-driven wearable biosensor technologies (including ECG monitors, continuous glucose monitors, respiratory sensors, edge-computing devices, explainable AI architectures, federated learning systems) |
| C — Comparator | Conventional (non-ML-integrated) wearable monitoring approaches or standard clinical monitoring; comparators are not systematically defined |
| O — Outcomes | Real-time physiological monitoring accuracy, early disease detection capability, precision medicine applicability, clinical translation feasibility, and patient outcomes (not formally quantified) |
Bottom Line
This paper presents itself as a comprehensive systematic review of ML-driven wearable sensors in healthcare but fails to meet the methodological standards of a systematic review by any established framework, including PRISMA, Cochrane, or CEBM criteria. There is no documented search strategy, no pre-specified inclusion criteria, no quality appraisal of included studies, and no quantitative synthesis of outcomes. The CEBM score of 10/100 reflects these fundamental methodological deficiencies. The review is more accurately characterised as a narrative overview or expert commentary on an emerging technology landscape. While it provides a useful conceptual map of ML-wearable sensor domains — ECG, glucose monitoring, respiratory sensing — and identifies genuine translational barriers including regulatory complexity, data standardisation, and algorithmic interpretability, it cannot be used to support evidence-based clinical decision-making. Senior clinicians and health technology assessment bodies should treat its conclusions as hypothesis-generating rather than practice-changing. For Australian practitioners, existing TGA SaMD frameworks and PBS-listed continuous glucose monitoring devices represent the current evidence-anchored starting point for wearable biosensor integration, pending higher-quality systematic evidence.
Key Findings
P Value: Not reported. No inferential statistics are presented.
Effect Size: Not reported. No pooled effect sizes are calculated or presented.
Primary Outcome: No formally defined primary outcome. The review narratively describes ML-wearable sensor performance across ECG monitoring, continuous glucose monitoring, and respiratory pattern detection, concluding that ML integration enhances real-time physiological monitoring and early disease detection capability.
Nnt Or Sensitivity: Not systematically reported. Individual study-level diagnostic accuracy metrics (sensitivity, specificity, AUC) may be referenced narratively within the review but are not pooled or summarised in a clinically actionable format.
Confidence Interval: Not reported. No confidence intervals are provided for any outcome.
Clinical Application
Clinical feasibility of ML-wearable integration is discussed as promising but contingent on resolution of substantial barriers: regulatory approval pathways, data interoperability standards, algorithmic transparency requirements, cybersecurity and privacy frameworks, and health workforce digital literacy. Edge-computing miniaturisation and federated learning are identified as enabling technologies but remain largely pre-clinical or early-implementation stage. In the Australian context, ML-driven wearable devices would require TGA approval under the Software as a Medical Device (SaMD) framework, aligned with the TGA's Digital Health guidance and the International Medical Device Regulators Forum (IMDRF) SaMD classification system. Continuous glucose monitors (e.g., Dexcom G6, FreeStyle Libre) are already PBS-listed for eligible patients with Type 1 diabetes (PBS item numbers introduced 2022), representing an existing clinical pathway for wearable biosensor integration. The Australian Digital Health Agency's national digital health strategy and My Health Record infrastructure provide a framework for data integration, though interoperability with ML platforms remains underdeveloped. RACGP guidelines on digital health and chronic disease management are relevant for primary care implementation. Privacy considerations are governed by the Privacy Act 1988 and Australian Privacy Principles, with additional obligations under the My Health Records Act 2012. The review's discussion of federated learning is particularly relevant to Australia's distributed healthcare system spanning urban and remote settings. Broadly applicable conceptually to any patient population requiring physiological monitoring, including those with cardiovascular disease (ECG monitoring), diabetes (continuous glucose monitoring), and respiratory conditions. However, the absence of population-specific efficacy data prevents confident application to defined clinical groups.
Abstract
The integration of machine learning (ML) with advanced wearable sensor technologies is revolutionizing healthcare by enabling real-time, intelligent monitoring of physiological parameters such as electrocardiogram (ECG), blood glucose, and respiratory patterns. This review systematically examines the transformative potential of ML-driven biosensors across three core domains: health monitoring, early disease detection, and precision medicine. Key technological advancements-including self-optimizing sensor networks, explainable AI (XAI) architectures, and edge-computing-enabled miniaturized devices-are critically evaluated. Despite rapid progress, the translation of these technologies into clinical practice faces significant challenges, such as data standardization, algorithmic interpretability, privacy concerns, and regulatory hurdles. This paper also discusses emerging trends, including federated learning, quantum machine learning, and neural interfaces, which hold promise for overcoming these barriers. By addressing these challenges and leveraging ongoing interdisciplinary collaborations, ML-enhanced wearable systems are poised to redefine personalized medicine and proactive healthcare delivery on a global scale.
References
- 1.Zhao, M., Liu, R., Jin, S., Ren, B., & Zhang, Q. (2026). From data to diagnosis: A comprehensive review of machine learning-driven wearable sensors in healthcare. Bioelectrochemistry, 109228. https://doi.org/10.1016/j.bioelechem.2026.109228
This content is for educational purposes for healthcare professionals only and does not constitute clinical advice. Clinical decisions should be based on individual patient assessment, current guidelines, and appropriate specialist consultation. Editorial Standards · Privacy Policy · Terms of Service