A machine learning driven computationally efficient horse shoe shaped antenna design for internet of medical things.
Authors: Rasool Khan Umhara, Sheikh Javaid A, Junaid Aqib, Ashraf Shazia, Balkhi Altaf A
Journal: PloS one
Summary
# Editorial Summary: Machine Learning-Optimised Horseshoe Antenna for Wearable Health Monitoring Wearable biomedical devices that transmit vital signs data to centralised monitoring platforms require efficient, compact antennas that perform reliably when worn on the body—a challenge that conventional antenna design methods struggle to address quickly. Researchers used regression-based machine learning algorithms trained on 1,080 electromagnetically simulated data points to predict the operating frequency, radiation efficiency and specific absorption rate (SAR) of a horseshoe-shaped patch antenna, dramatically reducing design iteration time compared to traditional electromagnetic modelling approaches. The optimised antenna design achieved resonance at 2.45 GHz with a 62.07% radiation efficiency, acceptable on-body SAR of 1.89 W/kg and 1.91 dBi peak gain—performance metrics suitable for body-worn telemetry systems across the 1.75–2.98 GHz bandwidth. This computational acceleration matters for equine and human healthcare professionals alike, as it enables faster prototype development of wearable monitoring systems that could track performance metrics in sport horses or detect early signs of illness through continuous vital sign transmission. The systematic comparison of ML algorithms provides a practical framework for engineers designing future generations of compact biomedical antennas, potentially bringing real-time remote health monitoring closer to cost-effective, widespread clinical application.
Read the full abstract on PubMed
Practical Takeaways
- •This paper is not relevant to equine practice; it concerns human biomedical wearable antenna design for medical monitoring systems
- •The research addresses engineering and telecommunications applications, not equine health, performance, or management
Key Findings
- •Horse shoe shaped patch antenna designed for wearable biomedical devices resonates at 2.45 GHz with SAR of 1.89 W/kg
- •Machine learning regression models successfully predicted antenna performance parameters (frequency, radiation efficiency, SAR) with computational efficiency
- •Antenna demonstrated 62.07% radiation efficiency and 1.91 dBi peak gain when mounted on human body