Wideband Circularly Polarized Conformal Antenna with Physics-Informed Neural Network Modeling for IoBNT Capsule...

Area of Science:

• Biomedical Engineering
• Artificial Intelligence
• Wireless Communication

Background:

• The Internet of Bio-Nano Things (IoBNT) integrates AI and biotech for intelligent in-body medical devices.
• Capsule endoscopy requires compact, reliable antennas for continuous diagnosis and monitoring.

Purpose of the Study:

• To propose a compact, wideband, circularly polarized conformal microstrip antenna for capsule endoscopy.
• To introduce a physics-informed neural network (PINN) for modeling thermal responses in biological tissues.

Main Methods:

• Designed and simulated a conformal microstrip antenna for a 10 mm capsule.
• Utilized a PINN framework to predict tissue thermal response based on the bioheat equation.
• Validated antenna performance in realistic tissue phantoms and assessed safety compliance.

Main Results:

• Achieved a wide impedance bandwidth (2.06-5.39 GHz) and stable matching in biological tissues.
• Demonstrated reliable circular polarization (2.31-3.14 GHz ARBW) for robust communication.
• PINN enabled fast, generalizable thermal prediction; antenna met safety (SAR) and communication range (1-3 m) requirements.

Conclusions:

• The integrated antenna and AI framework offers a scalable solution for intelligent, safe, and adaptive IoBNT-enabled capsule endoscopy.
• This technology advances continuous in-body diagnosis and monitoring capabilities.
• The study highlights the synergy between advanced antenna design and AI for next-generation medical devices.