Electromagnetic field confinement-driven multi-resonator THz biosensing platform with Au–Ag–graphene hybrid coatings and angle-invariant machine learning prediction for point-of-care breast cancer diagnostics
摘要
Breast cancer remains the leading cause of cancer-related mortality among women worldwide, with survival rates declining sharply from over 90% at early stages to only 28% in metastatic cases. Conventional screening methods, such as mammography and biomarker assays, suffer from reduced sensitivity in dense breast tissue and limited ability to detect low-abundance biomarkers (< 10 pg/mL) characteristic of early-stage disease.To address these challenges, we propose a multi-resonator terahertz biosensor incorporating Au–Ag–graphene hybrid coating architecture that enables enhanced electromagnetic field confinement and tunable surface conductivity.The sensor design integrates concentric resonators with strategic material placement: a central gold-coated circular resonator (2.7 μm diameter), a silver-coated square ring resonator (3.8–4.3 μm) and an outer graphene-coated circular ring (4.3–4.8 μm), fabricated on SiO₂ substrate.Electromagnetic simulations using COMSOL Multiphysics demonstrate exceptional performance, achieving a sensitivity of 500 GHz/RIU, a FOM of 7RIU−1 and detection limit of 0.3,enabling sub-femtomolar biomarker detection. Integration with machine learning yields 87% prediction accuracy for sensor response under variable incident angles (0–80°),with optimization achieving 91–100% reliability.Furthermore, 60 GHz frequency shift enables clear discrimination of breast cancer tissue, while GCP modulation offers remarkable tunability.These results highlight the potential of the proposed hybrid biosensor as point-of-care diagnostic platform for early-stage breast cancer detection.