Scholarlink

<p>This study presents a metasurface biosensor comprising a conjugated MXene–graphene architecture, augmented by machine‐learning optimization to achieve angle‐independent, multiplexed detection of malaria biomarkers. Finite‐element simulations performed in COMSOL Multiphysics reveal that, as the graphene chemical potential is varied within the 0.1–1 THz bandwidth, the sensor’s peak absorption coefficient increases from 0.193 to 1.094. The device attains an exceptional sensitivity of 500&#xa0;GHz per refractive‐index unit (RIU) and a figure of merit (FoM) of 22.727 RIU⁻<sup>1</sup> over a refractive‐index range of 1.373–1.402, demonstrating its utility for malaria biomarker quantification. Angular‐incidence analyses confirm stable absorption spectra under oblique excitation up to 40°, underscoring the design’s robustness. Integration of a one‐dimensional convolutional neural network (1D-CNN) enables predictive modeling with an R<sup>2</sup> of 1.00 across incident angles from 0° to 40°, thereby verifying angle‐independent performance. Furthermore, the sensor exhibits a tunable resonance frequency spanning an 80&#xa0;GHz range (1.118 THz to 1.108 THz), facilitating precise spectral discrimination for quantitative malaria 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

Machine Learning-Enhanced Terahertz Biosensor with Mxene-Graphene Conjugate for High-Sensitivity Malaria Detection

  • 期刊论文

  • 发表日期：

    2025年7月2日

查看全文
  • S. Selva Birunda,
  • Hariharan Subramani,
  • Arun Kumar U,
  • Taha Sheheryar

摘要

This study presents a metasurface biosensor comprising a conjugated MXene–graphene architecture, augmented by machine‐learning optimization to achieve angle‐independent, multiplexed detection of malaria biomarkers. Finite‐element simulations performed in COMSOL Multiphysics reveal that, as the graphene chemical potential is varied within the 0.1–1 THz bandwidth, the sensor’s peak absorption coefficient increases from 0.193 to 1.094. The device attains an exceptional sensitivity of 500 GHz per refractive‐index unit (RIU) and a figure of merit (FoM) of 22.727 RIU⁻1 over a refractive‐index range of 1.373–1.402, demonstrating its utility for malaria biomarker quantification. Angular‐incidence analyses confirm stable absorption spectra under oblique excitation up to 40°, underscoring the design’s robustness. Integration of a one‐dimensional convolutional neural network (1D-CNN) enables predictive modeling with an R2 of 1.00 across incident angles from 0° to 40°, thereby verifying angle‐independent performance. Furthermore, the sensor exhibits a tunable resonance frequency spanning an 80 GHz range (1.118 THz to 1.108 THz), facilitating precise spectral discrimination for quantitative malaria diagnostics.