Answers are grounded in this paper and cite their sources.
Diabetes mellitus is a major global health challenge characterized by persistently elevated blood glucose levels and the need for frequent glucose monitoring to support effective disease management. Conventional glucose monitoring methods generally rely on invasive or minimally invasive sampling, which can cause discomfort and reduce patient compliance during repeated measurements. Consequently, noninvasive and label-free sensing technologies capable of rapid and sensitive glucose detection are of increasing interest. Terahertz (THz) technology has emerged as a promising platform for next-generation glucose biosensing because of its sensitivity to changes in the dielectric properties of biological analytes and its potential for rapid, nondestructive detection. In this study, a compact THz biosensor based on cascaded triple-spiral resonators integrated with graphene, black phosphorus (BP), and vanadium dioxide (VO2) is proposed for enhanced glucose sensing. The hybrid integration of these active materials provides strong electromagnetic confinement, anisotropic field manipulation, and dynamically tunable resonance characteristics within the THz regime. The electromagnetic response of the proposed structure is investigated using the wave concept iterative process (WCIP), enabling efficient modeling of multilayer planar configurations incorporating isotropic and anisotropic conductive interfaces. Numerical validation against analytical solutions confirms the accuracy and stability of the developed WCIP framework. To further optimize the sensor, a deep neural network (DNN) model is developed to predict optimal operating conditions from the relevant material parameters, thereby reducing the need for exhaustive parametric simulations. The DNN achieves excellent predictive performance, with anR²value of 0.998, a mean absolute error of 0.260, and a mean squared error of 0.120. At a glucose concentration of500~mg/dL, the DNN-guided optimization increases the sensitivity from6.95~THz/RIUfor the preoptimization configuration to19.87~THz/RIUfor the optimized configuration, corresponding to an enhancement of 185.9%. Across the investigated glucose concentration range of 80–500~mg/dL, the optimized sensitivity remains high, ranging from 18.73 to19.87~THz/RIU. These results demonstrate that the proposed graphene-/BP-/VO2-integrated THz biosensor, combined with DNN-assisted optimization, can provide substantially enhanced glucose-sensing performance while reducing computational optimization requirements. The proposed approach therefore offers a promising foundation for rapid, label-free, and potentially noninvasive glucose monitoring, with significant potential for future wearable healthcare devices and intelligent biomedical sensing systems.
No takes yet. Share an insight, caveat, or question.
Dhivya et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: