G.N. Ramachandran Biology Seminar Series 2024: Beyond the Noise in Subdermal Wearables: Overcoming Artefacts in Real-Time Metabolite Monitoring

Subdermal wearable biosensors offer a minimally invasive solution for continuous, real-time monitoring of critical health biomarkers, such as glucose and lactate, directly from the interstitial fluid. These devices are particularly well-suited for ambulatory settings, where traditional monitoring methods are limited by movement-induced artefacts and physiological interferences. This study presents an advanced subdermal biosensor platform that leverages AI-driven algorithms for calibration and artefact reduction, enhancing data accuracy and reliability in real-world, prolonged-use scenarios.

 

Human participants were fitted with subdermal biosensors, enabling continuous, real-time recording of glucose and lactate levels. Data was wirelessly transmitted to a cloud-based monitoring system, where AI algorithms addressed common challenges, including artefact rejection, sensor drift, and time lag. The AI-enhanced platform demonstrated high data fidelity across varying activity levels, significantly improving the reliability of monitoring outcomes.

 

For glucose tracking, Clarke Error Grid analysis showed that 95% of the data points from the subdermal patches, when compared to capillary blood samples and commercially available Freestyle Libre 3 CGM sensors, fell within clinically acceptable Zones A and B. This confirms the accuracy of the biosensors for continuous glucose monitoring. In lactate measurement, Bland-Altman analysis revealed strong agreement with capillary blood measurements, underscoring the biosensors' potential for applications in exercise and metabolic assessments.

 

The results highlight the efficacy of AI-augmented subdermal wearable biosensors in delivering reliable, real-time, personalized health monitoring. With high fidelity in ambulatory conditions, this platform represents a significant advancement in wearable technology, offering a practical and robust solution for chronic disease management, metabolic health optimization, and proactive healthcare. This technology sets a new benchmark for continuous, real-time monitoring, paving the way for future innovations in wearable biosensing applications.