Scholarlink

<p>The development of non-invasive, continuous glucose monitoring systems remains a critical challenge in diabetes management, affecting millions of people worldwide. This paper presents a biosensor utilizing a hybrid metasurface architecture enhanced with multiple two-dimensional (2D) materials for high-sensitivity glucose detection. The proposed sensor design features an arrangement of resonators coated with graphene, borophene, molybdenum disulfide and germanene, mounted on a silicon dioxide substrate. Through numerical simulations using COMSOL Multiphysics, we demonstrate exceptional sensing performance with a sensitivity of 1000&#xa0;GHz/RIU, a quality factor of 9.859, and a figure of merit (FOM) of 15.625 RIU. The sensor exhibits excellent tunability through graphene chemical potential modulation, with transmittance varying from 97.917% to 52.699% across the investigated range. Angular dependency analysis reveals robust performance with transmittance decreasing from 52.699% to 16.314% as the incidence angle increases from 0° to 80°. Frequency range analysis shows optimal performance at 0.631 THz with a detection range spanning from 0.626 to 0.631 THz. Machine learning validation using Random Forest methodology achieves 90% R<sup>2</sup> accuracy for angle predictions and 100% accuracy for polynomial complexity analysis. These results establish the sensor's potential for practical implementation in next-generation diabetes management systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI-Driven Non-Invasive Glucose Sensing Using a Graphene–Metal Hybrid Terahertz Metasurface SPR Biosensor

  • 期刊论文

  • 发表日期：

    2025年6月2日

查看全文
  • R. Mahalakshmi,
  • Jacob Wekalao,
  • Amuthakkannan Rajakannu

摘要

The development of non-invasive, continuous glucose monitoring systems remains a critical challenge in diabetes management, affecting millions of people worldwide. This paper presents a biosensor utilizing a hybrid metasurface architecture enhanced with multiple two-dimensional (2D) materials for high-sensitivity glucose detection. The proposed sensor design features an arrangement of resonators coated with graphene, borophene, molybdenum disulfide and germanene, mounted on a silicon dioxide substrate. Through numerical simulations using COMSOL Multiphysics, we demonstrate exceptional sensing performance with a sensitivity of 1000 GHz/RIU, a quality factor of 9.859, and a figure of merit (FOM) of 15.625 RIU. The sensor exhibits excellent tunability through graphene chemical potential modulation, with transmittance varying from 97.917% to 52.699% across the investigated range. Angular dependency analysis reveals robust performance with transmittance decreasing from 52.699% to 16.314% as the incidence angle increases from 0° to 80°. Frequency range analysis shows optimal performance at 0.631 THz with a detection range spanning from 0.626 to 0.631 THz. Machine learning validation using Random Forest methodology achieves 90% R2 accuracy for angle predictions and 100% accuracy for polynomial complexity analysis. These results establish the sensor's potential for practical implementation in next-generation diabetes management systems.