JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Open Specy 1.0: Automated (Hyper)spectroscopy for Microplastics

  • 0Moore Institute for Plastic Pollution Research, Long Beach, California 90815, United States.
Analytical Chemistry +

|

August 11, 2025

|

August 11, 2025

View abstract on PubMed

Summary

This summary is machine-generated.

Open Specy 1.0 automates microplastic spectral analysis, significantly reducing processing time. New algorithms and machine learning classifiers enhance efficiency and accuracy in identifying microplastic pollution.

Area Of Science

  • Environmental Science
  • Analytical Chemistry
  • Spectroscopy

Background

  • Microplastic spectral analysis is a critical but time-consuming step in pollution studies.
  • Traditional methods require extensive manual processing, limiting throughput.
  • Automated and hyperspectral imaging techniques offer potential for increased efficiency.

Purpose Of The Study

  • To introduce Open Specy 1.0, an updated software for automated microplastic spectral analysis.
  • To present new algorithms for automated processing, including smoothing and particle compression.
  • To evaluate the performance of machine learning classifiers and a comprehensive spectral library.

Main Methods

  • Integration of automated processing algorithms (smoothing, particle compression) into Open Specy.
  • Development of two machine learning classifiers (logistic regression, k medoids) using a large spectral library (>40,000 spectra).
  • Evaluation of hyperspectral smoothing, particle identification, compression, and splitting configurations.

Main Results

  • Open Specy 1.0 achieved combined recovery rates of 50-150% for particle counts, identities, and sizes with CV < 40%.
  • Hyperspectral smoothing improved recovery to 96% (CV = 38%) compared to non-smoothed controls (86% recovery, CV = 83%).
  • Spectral compression for particles was >3x faster with similar accuracy and reduced variability compared to pixel-by-pixel processing.

Conclusions

  • Open Specy 1.0 significantly enhances the efficiency and accuracy of microplastic spectral analysis.
  • Automated techniques, particularly hyperspectral smoothing and spectral compression, are effective.
  • Further refinement of particle splitting algorithms and identification routines is needed to address challenges like false positives/negatives.