Rapid COVID-19 Detection Using MEMS-Based FTIR Spectrometers and Machine Learning

Strategic Insights -

Research has shown that the use of miniaturized MEMS-based Fourier-transform infrared (FTIR) spectrometers integrated with machine learning models can detect COVID-19 in a rapid and reagent-free manner. The study presented portable spectrometers that analyzed 363 nasopharyngeal swab samples stored in viral transport medium (VTM), achieving high diagnostic performance. The researchers also successfully implemented interval partial least squares discriminant analysis (iPLS-DA) for spectral data preprocessing and analysis. The findings of this research have significant implications for the development of real-time, point-of-care (POC) testing for COVID-19.

Key Takeaways:

  • The study demonstrated the use of MEMS-based FTIR spectrometers for COVID-19 detection, achieving 79% accuracy and 98% sensitivity in the mid-infrared (MIR) range and 80% accuracy in the dry sample model.
  • The researchers used machine learning models to analyze spectral data, with interval partial least squares discriminant analysis (iPLS-DA) achieving high diagnostic performance.
  • The study showed that the full measurement process, including sample handling, completes in under six minutes, making it suitable for real-time, point-of-care (POC) testing.
  • The research team obtained financial support from Ain Shams University and collaboration with Si-Ware Systems.
  • The study published in Scientific Reports included additional authors Mazen Erfan, Bassem Mortada, Mohamed Gaber, Shereen Saeed, Ghada Ismail, Ahmed ElShafei, MennaAllah S. Mohamed, Bassam Saadany, Yasser M. Sabry, and Diaa Khalil.

Statistics:

  • The study analyzed 363 nasopharyngeal swab samples stored in viral transport medium (VTM).
  • The MIR wet sample model achieved a diagnostic performance with 79% accuracy, 98% sensitivity, and an area under the curve (AUC) of 0.8.
  • The MIR dry sample model achieved an 80% accuracy and an AUC of 0.79.
  • The NIR model reached 66% accuracy and an AUC of 0.64.

Sources:

  • Rapid reagent free COVID19 detection using MEMS based FTIR spectroscopy and machine learning in NIR and MIR regions. Scientific Reports, 2025;15(1):35014.
  • Nature Portfolio, Heidelberger Platz 3, Berlin, 14197, Germany. (Nature Publishing Group - www.nature.com/; Scientific Reports - www.nature.com/srep/)
  • Ahmed Abdelkhalik, Si-Ware Systems, Khaled Ibn El-Waleed Street, Heliopolis, Cairo, Egypt.