An optical high-sensitivity terahertz metasurface SPR sensor with a MoS₂–graphene hybrid structure and machine learning for refractive-index-based brain tumor biomarker sensing
摘要
Early identification of brain tumors requires sensors capable of detecting subtle refractive index variations associated with pathological tissue changes in a label-free and non-invasive manner. Surface Plasmon Resonance (SPR) sensors utilize the surface plasmons excited by incident light to measure changes in the dielectric properties of thin layers placed adjacent to metal surfaces thus they are highly sensitive to dielectric property changes which makes them excellent platforms for biomedical diagnostics. In this work, a THz metasurface SPR sensor is designed and validated that incorporates a graphene base with gold (Au) and MoS₂ coated resonators arranged in a compact A-shaped metasurface configuration. The hybrid material used for the resonators, along with the resonator’s geometry, were both optimized to provide enhanced confinement and tunability of the electromagnetic fields within the terahertz frequency range. Simulation results indicate that the proposed sensor provides a maximum sensitivity of 2308 GHz/RIU with a stable FWHM of 0.111 THz and a peak FOM of 20.79 RIU− 1 across the refractive index range. A machine learning algorithm was applied as a data-driven analytical tool for modeling the relationships between sensor operating parameters and sensor responses, indicating correlation coefficients (R2) of unity for the simulated data sets. These results demonstrate that the proposed metasurface sensor offers a robust and high-performance design for RI based sensing. The findings presented are based on numerical validation and the demonstrated sensing parameters indicate that the design has a strong potential for future experimental prototype and early-stage brain tumor biomarker sensing contributing to SDG 3 (Good Health and Well-Being) through improved early diagnostic capability.