We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Reducing SpectralConfusion in Microplastic Analysis:A U‑Net Deep Learning Approach
Summary
Researchers developed a U-Net deep learning model to address spectral confusion between polyethylene and fatty acids in Raman spectroscopy-based microplastic detection, training the model on spectra from polystyrene, polyethylene, stearic acid, oleic acid, fatty acid mixtures, and polypropylene. The model achieved precise classification and, combined with binarization techniques, offered scalable qualitative and quantitative analysis of microplastics in complex environmental samples.
Among the various analytical techniques that have been proposed with the growing significance of microplastic detection, Raman spectroscopy is a powerful technique for detecting microplastics. However, the structural similarity in Raman spectra between fatty acids and polyethylene (PE) frequently causes misclassification by HQI-based methods, particularly when analyzing environmental samples containing mixed fatty acids. Herein, a U-net-based deep learning model was employed to precisely classify PE, stearic acid (SA), oleic acid (OA), mixtures of SA and OA, sodium dodecyl sulfate (SDS), and polypropylene based on their Raman spectra. Additionally, by incorporating a binarization technique commonly utilized in material chemistry, high scalability for both qualitative and quantitative analyses is provided. Consequently, the U-net model achieved accuracy improvements over the Pearson correlation coefficient of 2.05% to 11.09% for spectra with high signal-to-noise ratio (SNR) and 21.21% to 48.97% for spectra with nonaveraged spectra. Additionally, it demonstrated at least 36.69% higher accuracy compared to metrics such as Spearman correlation coefficient, cosine similarity, and Manhattan/Euclidean distance. This deep learning-based approach significantly reduces the confusion between PE and fatty acids observed in conventional Raman spectral analyses of microplastics, thereby demonstrating its potential applicability in microplastic standardization and analysis fields.
Sign in to start a discussion.
More Papers Like This
Reducing Spectral Confusion in Microplastic Analysis: A U-Net Deep Learning Approach
A common problem in microplastic detection using Raman spectroscopy is that fatty acids in environmental samples look chemically similar to polyethylene (a common plastic), causing misidentification. This study trained a deep learning model (U-Net architecture) to distinguish polyethylene from fatty acids and other organic compounds based on subtle spectral differences, achieving accurate classification. Better detection methods are foundational to all microplastic research, and this AI-assisted approach could reduce false positives in environmental monitoring.
Automatic classification of microplastics and natural organic matter mixtures using a deep learning model
Researchers developed a deep learning model using a convolutional neural network with spatial attention to classify microplastics mixed with natural organic matter from Raman spectra. The model achieved 99.54% accuracy compared to just 31.44% from conventional spectral library software, demonstrating that AI-based approaches can dramatically improve microplastic identification accuracy while reducing the need for time-intensive preprocessing steps.
Cascaded Improved Neural Network for the Reconstruction, Classification, and Unmixing of the Raman Spectra of Mixed Microplastics.
Researchers developed a cascaded neural network combining reconstruction, classification, and spectral unmixing to analyze mixed microplastic Raman spectra, achieving improved identification accuracy under complex environmental conditions where traditional preprocessing algorithms struggle with overlapping spectral peaks.
Identification of microplastics using a convolutional neural network based on micro-Raman spectroscopy
Researchers combined micro-Raman spectroscopy with a neural network to identify microplastics, achieving over 99% accuracy across 10 different plastic types. The system was also tested on real environmental samples and performed well at classifying unknown particles. This AI-powered approach could make microplastic identification faster and more reliable for environmental monitoring.
Quantitative analysis of microplastics in water environments based on Raman spectroscopy and convolutional neural network
Researchers developed a method combining Raman spectroscopy with a convolutional neural network to measure microplastic concentrations in water. The approach achieved high accuracy across six different sizes of polyethylene particles in five real-world water environments, outperforming other machine learning models and offering a practical tool for quantitative microplastic monitoring.