Scholarlink

<p>Metabolic disorders such as diabetes, hypertension, obesity, and lipid abnormalities (DHOL) demand early, non-invasive detection to mitigate health risks. This study introduces a gold-coated photonic crystal fiber (PCF) surface plasmon resonance (SPR) sensor optimized for the near-infrared (NIR) spectrum (700–2500 nm), leveraging deeper tissue penetration and reduced scattering. The sensor design employs a quasi-honeycomb air-hole configuration in fused silica, coated with a 50 nm gold layer, to enhance refractive index (RI) sensitivity through localized SPR. Finite element simulations in COMSOL Multiphysics demonstrate strong coupling between the HE11-like core mode and surface plasmon polaritons, achieving resonance shifts proportional to biomarker concentrations. Key biomarkers (glucose, angiotensin II, leptin, cholesterol) were detected with sensitivities exceeding 92% and specificities above 90%, validated via a confusion matrix. A deep neural network (DNN) was integrated to predict optical parameters (core loss, confinement loss) from simulation data, reducing computation time by 99.99% compared to COMSOL. The DNN achieved mean absolute errors below 0.10 for core power and confinement loss predictions, enabling real-time analysis. This work bridges plasmonics, photonic engineering, and machine learning, offering a rapid, highly sensitive tool for metabolic disorder monitoring.</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

Design and Numerical Analysis of a Gold-Coated Photonic Crystal Fiber Sensor for Metabolic Disorder Detection with Deep Learning Assistance

  • 期刊论文

  • 发表日期：

    2025年4月11日

查看全文
  • Sunil Sharma,
  • Sandip Das,
  • Chin-Shiuh Shieh,
  • Mong-Fong Horng,
  • Lokesh Tharani,
  • Sonal Sharma,
  • Prashant Sharma,
  • Prasun Chakrabarti,
  • Yashwant Singh Rawal

摘要

Metabolic disorders such as diabetes, hypertension, obesity, and lipid abnormalities (DHOL) demand early, non-invasive detection to mitigate health risks. This study introduces a gold-coated photonic crystal fiber (PCF) surface plasmon resonance (SPR) sensor optimized for the near-infrared (NIR) spectrum (700–2500 nm), leveraging deeper tissue penetration and reduced scattering. The sensor design employs a quasi-honeycomb air-hole configuration in fused silica, coated with a 50 nm gold layer, to enhance refractive index (RI) sensitivity through localized SPR. Finite element simulations in COMSOL Multiphysics demonstrate strong coupling between the HE11-like core mode and surface plasmon polaritons, achieving resonance shifts proportional to biomarker concentrations. Key biomarkers (glucose, angiotensin II, leptin, cholesterol) were detected with sensitivities exceeding 92% and specificities above 90%, validated via a confusion matrix. A deep neural network (DNN) was integrated to predict optical parameters (core loss, confinement loss) from simulation data, reducing computation time by 99.99% compared to COMSOL. The DNN achieved mean absolute errors below 0.10 for core power and confinement loss predictions, enabling real-time analysis. This work bridges plasmonics, photonic engineering, and machine learning, offering a rapid, highly sensitive tool for metabolic disorder monitoring.