Machine learning-assisted design and optimization of a hybrid BK7/Ag/WS2/GaN/ZrN/graphene optical SPR biosensor for label-free brain tumor-related refractive-index sensing

Brain tumors are among the most aggressive neurological malignancies, necessitating highly sensitive and label-free diagnostic platforms. This study proposes a surface plasmon resonance (SPR)-based refractive index (RI) sensor employing a BK7/Ag (Silver)/WS 2 (tungsten disulfide)/GaN (Gallium Nitride)/ZrN(zirconium nitride)/graphene multilayer architecture for brain tumor biomarker detection. The synergistic integration of plasmonic, dielectric, refractory plasmonic, and two-dimensional materials enhances electromagnetic field confinement and plasmon–analyte interactions, thereby improving sensing performance. The sensor was analysed using the Transfer Matrix Method (TMM) and validated through Finite Element Method (FEM) simulations. The optimized structure achieved near-unity absorption (>99.9%), a maximum angular sensitivity of 443.038°/RIU, a figure of merit of 118.777 RIU −1 , and a minimum detection limit of 0.006 RIU. Optimum layer thicknesses were identified as 32 nm (Ag), 1.9 nm (WS 2 ), 3.0 nm (GaN), 1.5 nm (ZrN), and 0.34 nm (graphene). A one-dimensional convolutional neural network (1D-CNN) surrogate model achieved an R 2 exceeding 0.99, enabling rapid sensor optimization for future label-free biomedical diagnostics.