Microinterventional in-sensor computing system for real-time metabolic health assessment.

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Biosensor Development

Background:

• Microneedle biosensors offer dynamic interstitial fluid monitoring but face challenges with motion artifacts and high energy demands for data processing.
• Existing biosensors struggle with signal instability and require significant power for wireless data transmission and analysis.

Purpose of the Study:

• To develop a novel self-anchoring microneedle biosensor system with integrated in-sensor computing.
• To overcome limitations of motion artifacts and energy consumption in current wearable biomarker monitoring devices.

Main Methods:

• A bio-inspired, starfish-like suction cup mechanism was employed for microneedle self-anchoring, enhancing signal stability.
• A lightweight deep learning algorithm (43 KB) was deployed on an embedded circuit for on-chip data processing and closed-loop feedback.
• The system was validated in a porcine model for continuous biochemical monitoring.

Main Results:

• The self-anchoring mechanism reduced signal fluctuations by 38-fold and increased signal intensity by up to 5.49-fold compared to planar devices.
• The embedded system achieved 98.68% diagnostic accuracy with a 120-hour battery life.
• Continuous biochemical dynamics were successfully captured in a porcine model, demonstrating real-world applicability.

Conclusions:

• The developed bio-mimetic interface and on-chip computing system significantly improve the reliability and efficiency of microneedle biosensors.
• This technology enables next-generation wearables for high-fidelity, real-time metabolic risk stratification in dynamic environments.
• The system's robust design and efficient processing pave the way for advanced in-home health monitoring solutions.