Beyond raw data: AI-driven biosensor fusion for enhancing athletic performance

Abstract

We have all heard a lot about the potential of data enabled by artificial intelligence (AI) to improve performance. This progress has seen the steady advancements of wearable biosensors capable of generating live data and feedback on an athlete's physiological state, allowing for smarter training and enabling consistent performance improvement. However, wearable biosensors often struggle with noise and signal interference caused by various factors, including muscle movements, sweat, and environmental conditions. This study proposes the Smart Performance Analysis and Real-time Tracking Algorithm (SPARTA) for enhancing athletic performance using AI techniques. SPARTA leverages AI algorithms to analyze real-time physiological data - including heart rate, oxygen saturation, skin conductance, and cortisol levels - enabling dynamic adjustments to training loads and recovery protocols. Experimental evaluations using the Biosensor-Student Health Fitness Dataset (n=500 input samples) demonstrated SPARTA’ s capability to achieve 91.34% accuracy in SpO 2 monitoring, 88.72% precision in skin conductance detection, 82.64% correlation with laboratory assays for sweat electrolyte analysis, and 78.65% accuracy in non-invasive cortisol level tracking. With more advances in artificial intelligence, wearable biosensors will greatly help boost athletic performance, further dominating the sports & fitness globe.

Ähnliche Arbeiten