Advanced plasmonic sensor design for sperm detection with machine learning-driven optimization

Abstract

This study investigates the optimization of surface plasmon resonance parameters for the development of a high-sensitivity, label-free sperm detection system. An analysis was performed to evaluate critical optical and plasmonic parameters within the frequency range of 0.12–0.129 THz. A detailed parametric analysis was conducted, examining key sensing metrics including refractive index variations (n = 1.33–1.3461), sensitivity (S), figure of merit (FOM), detection limit (DL), and signal-to-noise ratio (SNR). Results demonstrate a significant enhancement in sensing performance, with optimal FOM values of 131.579RIU⁻¹ achieved at frequencies of 0.123 and 0.12 THz, and corresponding detection limit values reaching 0.013. The system exhibits stable spectral characteristics, maintaining a consistent full width at half maximum (FWHM) of 0.038 across the investigated frequency range. Peak sensitivity values of 5000 GHz/RIU were observed, concurrent with a constant detection angle (DA) of 26.316. The quality factor (Q) exhibited a decrease from 3.395 to 3.158 with decreasing frequency, indicating an optimal operational range. A controlled decrease in the dynamic range (DR) from 0.662 to 0.616 was also observed. The optimized parameters demonstrate the potential for implementing highly sensitive, reliable SPR platforms suitable for both clinical diagnostics and research applications in biomedical applications.