Detection of Proteins in a Surface Plasmon Resonance Biosensor Based on Hybrid Metasurface Architecture and Behaviour Prediction Using Random Forest Regression

The emergent paradigm of nanoscale biosensing technologies represents a transformative modality in biomolecular detection, with substantive implications across critical interdisciplinary domains including clinical diagnostics, environmental monitoring, and alimentary safety assessment. This investigation presents a protein detection biosensor predicated on a simple hybrid metasurface architecture, integrating methylammonium lead halide perovskite and graphene nanomaterials through a quantum-engineered approach. The proposed sensor architecture employs a dual-resonator configuration with orthogonal geometric morphologies—circular and square geometries—optimized through first-principles computational simulations to maximize electromagnetic coupling and enhance biochemical sensing transduction. Using high-resolution finite element modelling and spectroscopic analysis, we systematically characterize the sensor’s intrinsic capacity for detecting nanomolar protein concentration gradients and refractive index modulations. Spectral characterization reveals exceptional sensitivity metrics, demonstrating resonance frequency shifts with sensitivity 161–226 GHzRIU⁻¹. The proposed design also demonstrates excellent detection limit of 0.476 to 0.719 which is indicative of its quantum-level sensing capabilities. Additionally, the integration of Random Forest Regression demonstrates remarkable performance with optimal R² score of 100% across all the cases. The results confirm the sensor’s precision, highlighting its applicability in biomolecular detection.