Highly Responsive Graphene-Integrated Metasurface Surface Plasmon Resonance Sensor Augmented with Ag-Au-SiO₂ Nanostructures and Machine Learning for Advanced Biomedical Diagnostics
摘要
This study report the development of a high-performance terahertz biosensor for non-invasive detection and assessment of sperm quality, offering an advanced approach to reproductive health diagnostics. The device employs a multilayered metasurface architecture composed of copper circular resonators, a silver-enhanced square element, and a gold rectangular ring, all integrated on a graphene-coated substrate. This configuration enables strong field confinement and enhanced sensitivity within the 0.177–0.186 THz frequency band. The sensor achieves a peak sensitivity of 5000 GHz/RIU, with quality factors between 4.214 and 4.429, a signal-to-noise ratio of 0.024, and a detection accuracy of 23.810. Integration with artificial intelligence algorithms significantly enhances predictive performance, yielding R2 values of 1.0 across incident angles from 0° to 80° and refractive index changes from 1.33 to 1.3461. This AI-augmented terahertz platform represents a robust and rapid tool for sperm analysis, addressing critical needs in male fertility evaluation without reliance on invasive procedures.