Graphene Metasurface-Based Biosensor for Direct Dopamine Detection Utilizing Surface Plasmon Resonance in the Terahertz Regime with Machine Learning Optimization via K-Nearest Neighbors Regression

This study introduces an innovative terahertz-based biosensor designed for the direct detection of dopamine. By leveraging surface plasmon resonance (SPR) principles, extensive computational electromagnetic simulations are conducted to optimize the sensor design across the 0.1-0.6 THz frequency spectrum. The proposed biosensor exhibits exceptional sensitivity to dopamine concentrations ranging from 10 to 0.001 ppm, corresponding to refractive indices between 1.256 and 1.309 RIU. The sensor's performance is characterized by impressive metrics, including a peak sensitivity of 500 GHzRIU-1, a figure of merit of 2.809 RIU(-)1, and a detection limit of 0. 867RIU.To further enhance the sensor's predictive capabilities, we have implemented a K-nearest neighbors (KNN) regression model. This machine learning approach is employed to forecast absorption values based on various structural parameters. The model demonstrates remarkable accuracy, achieving R2 scores up to 1.0 across diverse test cases and K values. The proposed biosensor outperforms many existing sensors in terms of sensitivity and detection limit, showing significant potential for early diagnosis and monitoring of dopamine-related neurological conditions. By integrating advanced materials, innovative design, and machine learning techniques, the proposed approach represents a significant advancement in dopamine detection methodologies for clinical applications. The proposed biosensor demonstrates the potential to revolutionize the diagnosis and treatment of disorders such as Parkinson's disease, schizophrenia, and attention deficit hyperactivity disorder (ADHD).