Scholarlink

<p>This work presents an innovative X-shaped photonic crystal fiber (PCF) sensor based on surface plasmon resonance (SPR), specifically designed for the sensitive and accurate detection of malaria through blood sample analysis. The optical and structural features of the device were investigated via the finite element method (FEM) incorporating perfectly matched layer (PML) boundaries to guarantee accurate modal analysis. To enhance detection performance, the device’s geometry and material parameters were fine-tuned using the Box-Behnken Design (BBD) methodology. The optimized configuration achieved remarkable sensitivity to refractive index changes associated with different malaria stages: 2857.14&#xa0;nm/RIU and 38.05 RIU<sup>− 1</sup> for the ring stage, 2105.26&#xa0;nm/RIU and 120.30 RIU<sup>− 1</sup> for the trophozoite stage, and 2068.97&#xa0;nm/RIU and 207.19 RIU<sup>− 1</sup> for the schizont stage. These findings highlight the biosensor’s capability for both early detection and stage differentiation of malaria. Furthermore, machine learning regression algorithms were employed to model and forecast confinement loss (CL), with the Random Forest Regressor (RFR) delivering the best predictive performance. The results emphasize the synergistic use of FEM simulations, experimental design strategies, and machine learning techniques in advancing PCF-SPR biosensors for real-time, high-sensitivity biomedical 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

Hybrid BBD-ML optimized X-shaped PCF-SPR sensor for sensitive malaria detection

  • 期刊论文

  • 发表日期：

    2025年12月20日

查看全文
  • Sameh Kaziz,
  • Lamia Guedri-Knani,
  • Chérif Dridi

摘要

This work presents an innovative X-shaped photonic crystal fiber (PCF) sensor based on surface plasmon resonance (SPR), specifically designed for the sensitive and accurate detection of malaria through blood sample analysis. The optical and structural features of the device were investigated via the finite element method (FEM) incorporating perfectly matched layer (PML) boundaries to guarantee accurate modal analysis. To enhance detection performance, the device’s geometry and material parameters were fine-tuned using the Box-Behnken Design (BBD) methodology. The optimized configuration achieved remarkable sensitivity to refractive index changes associated with different malaria stages: 2857.14 nm/RIU and 38.05 RIU− 1 for the ring stage, 2105.26 nm/RIU and 120.30 RIU− 1 for the trophozoite stage, and 2068.97 nm/RIU and 207.19 RIU− 1 for the schizont stage. These findings highlight the biosensor’s capability for both early detection and stage differentiation of malaria. Furthermore, machine learning regression algorithms were employed to model and forecast confinement loss (CL), with the Random Forest Regressor (RFR) delivering the best predictive performance. The results emphasize the synergistic use of FEM simulations, experimental design strategies, and machine learning techniques in advancing PCF-SPR biosensors for real-time, high-sensitivity biomedical diagnostics.