Scholarlink

<p>This study reports a biosensor engineered for the label-free detection of breast cancer biomarkers, leveraging surface plasmon polaritons in a metal–insulator–metal configuration. The sensor architecture integrates gold-coated rectangular resonators and silver-coated square resonators atop a graphene-functionalized circular base further enhanced by a surrounding array of MXene-coated hemispherical ring resonators on a SiO₂ substrate. Finite element simulations using COMSOL Multiphysics demonstrates a resonant frequency window between 0.738 and 0.742 THz, delivering a sensitivity of 500 GHzRIU<sup>−1</sup>a figure of merit (FOM) of 5.208 RIU⁻<sup>1</sup>, a quality factor (Q) of 7.729, and a detection limit of 0.401 RIU. The sensor exhibits a strong linear dependence of resonance frequency on refractive index (R<sup>2</sup> = 0.988), and design optimization guided by machine learning—specifically polynomial regression—achieves predictive accuracies approaching 100%. These findings position the proposed device among the most sensitive THz biosensors and underscore its promise for non-invasive, real-time breast cancer diagnostics.</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

Artificial Intelligence-Enhanced Terahertz Metasurface Biosensor for Breast Cancer Biomarker Detection

  • 期刊论文

  • 发表日期：

    2025年5月30日

查看全文
  • Jothi Prabha Appadurai,
  • Kumaravel Kaliaperumal,
  • Jacob Wekalao,
  • Amuthakkannan Rajakannu

摘要

This study reports a biosensor engineered for the label-free detection of breast cancer biomarkers, leveraging surface plasmon polaritons in a metal–insulator–metal configuration. The sensor architecture integrates gold-coated rectangular resonators and silver-coated square resonators atop a graphene-functionalized circular base further enhanced by a surrounding array of MXene-coated hemispherical ring resonators on a SiO₂ substrate. Finite element simulations using COMSOL Multiphysics demonstrates a resonant frequency window between 0.738 and 0.742 THz, delivering a sensitivity of 500 GHzRIU−1a figure of merit (FOM) of 5.208 RIU⁻1, a quality factor (Q) of 7.729, and a detection limit of 0.401 RIU. The sensor exhibits a strong linear dependence of resonance frequency on refractive index (R2 = 0.988), and design optimization guided by machine learning—specifically polynomial regression—achieves predictive accuracies approaching 100%. These findings position the proposed device among the most sensitive THz biosensors and underscore its promise for non-invasive, real-time breast cancer diagnostics.