EUVIMEDOntdek EUVIMED
Open access · CC BY-NC-ND 4.0

Enhancing Confidence in Microplastic Spectral Identification via Conformal Prediction

Madeline E Clough iD, Eduardo Ochoa Rivera, Rebecca L Parham iD, Andrew P Ault iD, Paul M Zimmerman iD, Anne J McNeil iD, Ambuj Tewari iD

Environmental Science & Technology · 2024-12-10 · 58 (49) · 21740-21749

DOI: 10.1021/acs.est.4c05167 ↗

At a glance

Journal article in Environmental Science & Technology, published online on 2024-11-26. Written by Madeline E Clough and 6 co-authors. Institution: Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109-1055, United States. Major MEDLINE topics: Microplastics. Cited 29 times so far (Crossref); 60 references. The full text is freely available (open access, CC BY-NC-ND 4.0). Funded by: College of Literature, Science, and the Arts, Directorate for Mathematical and Physical Sciences.

SamenvattingSubjectsBibliografische gegevensCiteRelated studies

Abstract (original language)

Microplastics are an emerging pollutant of concern, with environmental observations recorded across the world. Identifying the type of microplastic is challenging due to spectral similarities among the most common polymers, necessitating methods that can confidently distinguish plastic identities. In practice, a researcher chooses the reference vibrational spectrum that is most like the unknown spectrum, where the likeness between the two spectra is expressed numerically as the hit quality index (HQI). Despite the widespread use of HQI thresholds in the literature, acceptance of a spectral label often lacks any associated confidence. To address this gap, we apply a machine-learning framework called conformal prediction to output a set of possible labels that contain the true identity of the unknown spectrum with a user-defined probability (e.g., 90%). Microplastic reference libraries of environmentally aged and pristine polymeric materials, as well as unknown environmental plastic spectra, were employed to illustrate the benefits of this approach when used with two similarity metrics to compute HQI. We present an adaptable workflow using our open-access code to ensure spectral matching confidence for the microplastic community, reducing manual inspection of spectral matches and enhancing the robustness of quantification in the field.

Subjects

MeSH terms (MEDLINE)

MicroplasticsEnvironmental MonitoringMachine LearningPlastics

Author keywords

conformal predictionhit quality indexmicroplasticsspectral librariesspectral matching

Substances

MicroplasticsPlastics

Authors and institutions

  • Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109-1055, United States.
  • Department of Statistics, University of Michigan, 1085 South University Avenue, Ann Arbor, Michigan 48109-1055, United States.
  • Macromolecular Science and Engineering Program, University of Michigan, 2300 Hayward Street, Ann Arbor, Michigan 48109-1055, United States.
  • Department of Electrical Engineering and Computer Science, 1301 Beal Avenue, University of Michigan, Ann Arbor, Michigan 48109-1055, United States.

ORCID profiles: Madeline E Clough, Rebecca L Parham, Andrew P Ault, Paul M Zimmerman, Anne J McNeil, Ambuj Tewari

Funding

  • College of Literature, Science, and the Arts
  • Directorate for Mathematical and Physical Sciences

Cite

APA

Clough, M. E., Ochoa Rivera, E., Parham, R. L., Ault, A. P., Zimmerman, P. M., McNeil, A. J., & Tewari, A. (2024). Enhancing Confidence in Microplastic Spectral Identification via Conformal Prediction. Environmental Science & Technology, 58(49), 21740-21749. https://doi.org/10.1021/acs.est.4c05167

Vancouver

Clough ME, Ochoa Rivera E, Parham RL, Ault AP, Zimmerman PM, McNeil AJ, et al. Enhancing Confidence in Microplastic Spectral Identification via Conformal Prediction. Environmental Science & Technology. 2024;58(49):21740-21749. doi:10.1021/acs.est.4c05167

BibTeX
@article{Clough2024,
  author = {Clough, Madeline E and Ochoa Rivera, Eduardo and Parham, Rebecca L and Ault, Andrew P and Zimmerman, Paul M and McNeil, Anne J and Tewari, Ambuj},
  title = {Enhancing Confidence in Microplastic Spectral Identification via Conformal Prediction},
  journal = {Environmental Science & Technology},
  year = {2024},
  volume = {58},
  number = {49},
  pages = {21740--21749},
  doi = {10.1021/acs.est.4c05167}
}
Start for free →