NUS Sensor Tracks Fatigue, Stress on the Go: A Revolutionary Step Towards Real-World Mental Health Monitoring
In today's fast-paced world, the signs of burnout and chronic fatigue are all too familiar. According to recent studies, approximately one in three employees in Singapore report feeling burnt out, a statistic that is alarmingly high and indicative of a growing global concern. The economic and health implications of these conditions are significant, especially in professions where alertness is paramount. However, diagnosing fatigue and related mental health issues has traditionally relied on self-reported questionnaires, which are inherently subjective and often insufficient for real-time evaluation.
This is where wearable technology steps in, offering a potential solution by continuously tracking cardiovascular markers linked to the autonomic nervous system. Yet, the challenge lies in the fact that these devices struggle to capture accurate signals during everyday movement due to motion artefacts from muscle activity, body movement, and physiological interference. Current mitigation strategies often fail to address the complexity of these noise sources, resulting in poor signal quality.
A groundbreaking research team led by Professor Ho Ghim Wei from the Department of Electrical and Computer Engineering at the National University of Singapore has developed a novel solution. They introduced a metahydrogel platform integrated with AI-driven signal processing, which simultaneously suppresses multiple sources of motion noise. This innovation has achieved remarkable accuracy in tracking fatigue levels, meeting ISO clinical-grade standards, and surpassing commercial trackers in the market.
The metahydrogel artefact-mitigating platform (MAP) is a marvel of engineering. It employs two filtering mechanisms within a single material. Nanoparticles self-assemble into periodic bands, scattering and absorbing mechanical vibrations akin to soundproofing panels. This mechanism blocks movement noise within specific frequency ranges. Simultaneously, a biocompatible glycerol-water electrolyte controls ion travel, allowing low-frequency heart signals to pass through while suppressing higher-frequency muscle electrical noise. A machine-learning denoising algorithm further refines the signal, ensuring the preservation of critical physiological features.
The platform's design is not just about improved hardware; it's also about matching the mechanical properties of biological tissue. It is soft, breathable, and durable under repeated stretching, making it highly compatible with the human body. By combining enhanced hardware with smart algorithms, the system significantly boosts signal quality, making it easier to detect fatigue-related patterns in heart rhythms. This advancement increases peak-detection accuracy from 52% to 93%, a substantial improvement over current commercial devices.
Dr. Tian, a key member of the research team, highlights the platform's superiority, especially under motion conditions. Current smartwatches, for instance, typically achieve ECG signal-to-noise ratios of 10-20 dB, which can drop by approximately 40% during movement. In contrast, the NUS sensor achieves a remarkable 37 dB during daily activities, showcasing its superior performance.
The research team's wearable MAP system, with wireless transmission capabilities, has been successfully tested in real-world scenarios, including simulated driving tasks designed to induce fatigue. High-quality cardiovascular data collected from the hydrogel sensor has enabled a deep-learning system to identify fatigue levels with 92% accuracy, a significant improvement over the 64% accuracy when trained on data without MAP. Moreover, the system meets the ISO 81060-2 gold-standard requirements for blood pressure monitoring, further solidifying its reliability.
The implications of this research extend beyond fatigue tracking. MAP has demonstrated its ability to suppress artefacts across various biosignal types, including heart sounds, respiratory sounds, voice, brain waves, and eye movements. This versatility opens up exciting possibilities for broader neurophysiological and mental health monitoring.
The journey to this breakthrough began with a four-year focus on developing the underlying sensing technologies. The team then transitioned to the metahydrogel concept, which emerged about two and a half years ago. The platform's design and fabrication took approximately a year, during which the researchers crafted a library of metahydrogels with different material systems to target noise across various frequency ranges. Subsequent stages involved system integration and application validation, including exploring its potential for mental health monitoring.
Looking ahead, the team aims to collaborate with mental health physicians to understand the most relevant physiological data in real-world settings and the accuracy levels required to meet clinical needs. They also seek industrial partners to optimize manufacturing strategies and advance the platform towards practical, product-level implementation. This collaboration is crucial to translating laboratory-based research into real-world applications.
In conclusion, the NUS sensor's ability to track fatigue and stress on the go is a significant step forward in real-world mental health monitoring. It offers a promising solution to the challenges posed by burnout and chronic fatigue, potentially transforming the way we approach mental health assessment and management.
