Design and Investigation of Machine Learning–Optimized Surface Plasmon Resonance Biosensor for Early Brain Tumor Detection

Brain tumors are indeed a critical health concern, and timely detection plays a pivotal role in improving patient outcomes. Properly executed brain cancer recognition methodologies can save valuable lives. These techniques require remarkable mobility, exactness, prompt reaction speed, and superior sensitivity. Brain tumor sensors possess the potential to greatly enhance the prospects of patients by facilitating swift identification and therapy. Detecting brain cancers early using these sensors allows for timely intervention and better sufferer results. This research represents a significant step toward achieving these goals. Variations in different materials (bismuth, gold, and silver) are investigated as a resonator to get the best response characteristic. In addition, machine learning prediction is employed for in-depth analysis and parameter improvement. As a result, the amount of time required to collect data was significantly reduced as compared to running simulations, which need a step size of around 6 h for each variation. The proposed novel T-shaped square SSR-based biosensor uniqueness lies in its ability to identify and distinguish between diverse brain tissues. This novel T-shape allows for label-free and real-time monitoring of the brain tumor cells. This innovative geometry structure gives early and accurate detection that can potentially improve patient outcomes. It achieves the highest sensitivity of 850.34 nm/RIU for metastasis, accompanied by a figure of merit (FOM) of 7.541/RIU for multi-sclerosis determination and an impressive quality factor of 14.29 for a wall of solid brain. Operating within the wavelength span of 1200 nm to 2300 nm, this sensor holds great promise for efficient and accurate brain tissue differentiation.