Enhanced Graphene-Based Metasurface Biosensor for Brain Tumor Detection and Behavior Prediction Using Random Forest Regression

This manuscript introduces a terahertz biosensor that integrates graphene-enhanced metasurfaces with a dual-resonator configuration for the accurate identification of brain cancers. The sensor comprises concentric circular resonators encircled by rectangular components on a silicon dioxide substrate, using graphene’s adjustable chemical potential to improve its efficacy. A thorough analysis of the sensor’s geometric variables is conducted to optimize its design. The results show an optimum sensitivity of 769 GHz RIU −1 and a figure of merit of 28.49 RIU −1 , demonstrating its potential for accurate biomarker detection. Additionally, random forest regression is used to model and predict the sensor’s response under different biochemical conditions, integrating machine learning with the metasurface platform to improve detection accuracy and provide a robust framework for real-time diagnostics.