Scholarlink

<p>This study evaluates the potential of Electrical Impedance Spectroscopy (EIS) combined with machine learning to achieve accurate, non-destructive assessment of meat freshness. Twenty-four Holstein beef samples were monitored over an eight-day storage period, with physicochemical parameters (color, pH, dry matter, mineral and protein content) and impedance measurements collected across 1–10⁵ Hz. The impedance marker log|Z″| showed a strong correlation with storage duration (R² = 0.83), demonstrating its predictive value for quality degradation. Random forest model without data transformation achieved near-perfect classification of fresh versus stale meat (AUC ≈ 1.0). These findings highlight the promise of integrating EIS with machine learning for real-time, reliable freshness detection and provide a foundation for smart impedimetric sensing technologies in the meat industry.</p> Graphical Abstract <p></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

Random forest–based electrical impedance spectroscopy for precise freshness classification of Holstein beef

  • 期刊论文

  • 发表日期：

    2026年2月14日

查看全文
  • Chaima Dhifallah,
  • Sami Ameur,
  • Amine Mosbah,
  • Lotfi Mhamdi,
  • Naceur M’Hamdi,
  • Amine Mezni,
  • Ridha Ajjel

摘要

This study evaluates the potential of Electrical Impedance Spectroscopy (EIS) combined with machine learning to achieve accurate, non-destructive assessment of meat freshness. Twenty-four Holstein beef samples were monitored over an eight-day storage period, with physicochemical parameters (color, pH, dry matter, mineral and protein content) and impedance measurements collected across 1–10⁵ Hz. The impedance marker log|Z″| showed a strong correlation with storage duration (R² = 0.83), demonstrating its predictive value for quality degradation. Random forest model without data transformation achieved near-perfect classification of fresh versus stale meat (AUC ≈ 1.0). These findings highlight the promise of integrating EIS with machine learning for real-time, reliable freshness detection and provide a foundation for smart impedimetric sensing technologies in the meat industry.

Graphical Abstract