Customizable Machine-Learning Models for Rapid Microplastic Identification Using Raman Microscopy

Abstract

Raman spectroscopy is commonly used in microplastics identification, but equipment variations yield inconsistent data structures that disrupt the development of communal analytical tools. We report a strategy to overcome the issue using a database of high-resolution, full-window Raman spectra. This approach enables customizable analytical tools to be easily createda feature we demonstrate by creating machine-learning classification models using open-source random-forest, K-nearest neighbors, and multi-layer perceptron algorithms. These models yield >95% classification accuracy when trained on spectroscopic data with spectroscopic data downgraded to 1, 2, 4, or 8 cm–1 spacings in Raman shift. The accuracy can be maintained even in non-ideal conditions, such as with spectroscopic sampling rates of 1 kHz and when microplastic particles are outside the focal plane of the laser. This approach enables the creation of classification models that are robust and adaptable to varied spectrometer setups and experimental needs.

Saccharide Concentration Prediction from Proxy Ocean Samples Analyzed Via Infrared Spectroscopy and Quantitative Machine Learning

“…SVRs have also been employed in the classification of Raman spectra to identify Alzheimer’s disease in mice; a relevant features map is utilized to identify pertinent peaks that are from molecules known to be associated with the disease. A study from 2022 reports comparable classification accuracy of microplastic Raman microscopy samples from k-nearest neighbors (KNN), multilayer perceptron (MLP), and random forest (RF) models . These literature examples highlight the diverse applications of ML and develop techniques that expand the applications of chemistry, as we present herein.…”

“…A study from 2022 reports comparable classification accuracy of microplastic Raman microscopy samples from k-nearest neighbors (KNN), multilayer perceptron (MLP), and random forest (RF) models. 49 These literature examples highlight the diverse applications of ML and develop techniques that expand the applications of chemistry, as we present herein.…”

When microplastics meet electroanalysis: future analytical trends for an emerging threat

“…An advantage of their work is the ability to provide a particle analysis describing the number of particles and some other characteristics (such as diameter) for each polymer type. 154 Regarding AI in m-Raman, Lei et al 155 presented a proof-ofconcept ML tool KNN with the aim of reducing the inconsistencies in the detection and identication of MPs by m-Raman. The authors tested a training model for ML with 4520 spectra, using 40 reference spectra and 108 samples of 14 types of commercial MPs (PE, PP, PTFE, PS, PU, PMMA, PET, PVC, PC, ABS, PES, polyoxymethylene (POM), PS and PA), and found a spectral agreement of more than 95% at a rate of 10 samples per s.…”

Saccharide concentration prediction from proxy-sea surface microlayer samples analyzed via ATR-FTIR spectroscopy and quantitative machine learning

“…A study from 2022 reports comparable classification accuracy of microplastic Raman microscopy samples from k-nearest neighbors (KNN), multi-layer perceptron (MLP), and random forest (RF) models. 41 These literature examples highlight the diverse applications of ML and develop techniques that expand the applications of chemistry, as we present herein.…”