Machine learning-integrated surface-enhanced Raman spectroscopy analysis of multicomponent dye mixtures

Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy, Journal Year: 2025, Volume and Issue: unknown, P. 125806 - 125806

Published: Jan. 1, 2025

Language: Английский

---

Automatic microplastic classification using dual-modality spectral and image data for enhanced accuracy

Marine Pollution Bulletin, Journal Year: 2025, Volume and Issue: 213, P. 117665 - 117665

Published: Feb. 17, 2025

Language: Английский

---

Recent advancements in SERS-based detection of micro- and nanoplastics in food and beverages: techniques, instruments, and machine learning integration

Trends in Food Science & Technology, Journal Year: 2025, Volume and Issue: unknown, P. 104940 - 104940

Published: Feb. 1, 2025

Language: Английский

---

Identification of marine microplastics by a combined method of principal component analysis and random forest for fluorescence spectrum processing

Marine Pollution Bulletin, Journal Year: 2025, Volume and Issue: 214, P. 117740 - 117740

Published: Feb. 26, 2025

Language: Английский

---

Biosensors for Micro- and Nanoplastics Detection: A Review

Chemosensors, Journal Year: 2025, Volume and Issue: 13(4), P. 143 - 143

Published: April 14, 2025

Microplastics (MPs) and nanoplastics (NPs), which are widespread in many habitats as the byproducts of various industrial processes, pose considerable environmental health hazards. However, current, conventional methods for detecting characterizing them considerably lacking throughput, sensitivity, reliability, field deployability. In current report, we review state art biosensor-based MP/NP detection, particular, describing advances optical electrochemical approaches, along with development novel biorecognition elements application bioinformatics tools.

Language: Английский

---

SpecRecFormer: Deep Learning-Driven Adaptive Component Identification of PAH Mixtures Based on Single-Component Raman Spectra

Analytical Chemistry, Journal Year: 2025, Volume and Issue: unknown

Published: April 29, 2025

The identification of components in mixed spectra is a fundamental challenge spectral analysis, complicated by factors such as peak overlap due to structural similarities, shifts characteristic peaks from molecular interactions, and interferences caused matrix effects. While deep learning offers robust feature extraction capabilities notable advantages addressing these challenges, it still faces significant obstacles, including the limited availability labeled data for effective training difficulty applying fixed-threshold predictive models containing uncertain components. This paper established model, SpecRecFormer, rapid individual polycyclic aromatic hydrocarbons (PAHs) based on their Raman spectra. model integrates dual-channel convolutional neural network (CNN) local with Transformer module global representation. It trained reference database composed single-component spectra, simulated generated through augmentation expand diversify set. architecture enables evaluate similarity between unknown known references. To further enhance recognition accuracy, an adaptive threshold strategy introduced, dynamically adjusting decision thresholds characteristics retain only exceeding candidate predictions. Experimental results demonstrate that derived four generalizes effectively three real-world PAH sets, achieving accuracies 93.75%, 89.21%, 93.63%, respectively, significantly outperforming conventional models. These findings present innovative highly approach substantial potential advancing applications environmental science chemical analysis.

Language: Английский

---

Supervised classification combined with genetic algorithm variable selection for a fast identification of polymeric microdebris using infrared reflectance

Marine Pollution Bulletin, Journal Year: 2023, Volume and Issue: 195, P. 115540 - 115540

Published: Sept. 16, 2023

Pollution caused by plastics and, in particular, microplastics has become a source of environmental concern for Society. Their ubiquity, with millions tons plastic debris spilled both land and sea, requires efficient technological improvements the ways residues are collected, handled, characterized recycled. For reliable decision-making, dependable chemical information is essential to assess nature found environment their fate. In this work an method identify polymeric composition microplastic fragments proposed. It combines infrared reflectance spectra chemometric methods. A breakthrough result that models include polymers weathered under dry (shoreline) submerged (in sea water) conditions hence, they very promising as starting point eventual practical applications. addition, no spectral processing required after initial measurement. SYNOPSIS: This approach aquatic environments measurements multivariate data analysis fight against (micro)plastic pollution.

Language: Английский

---

A Study of High-Frequency Noise for Microplastics Classification Using Raman Spectroscopy and Machine Learning

Applied Spectroscopy, Journal Year: 2024, Volume and Issue: 78(6), P. 567 - 578

Published: March 11, 2024

Given the growing urge for plastic management and regulation in world, recent studies have investigated problem of material identification correct classification disposal. Recent works shown potential machine learning techniques successful microplastics using Raman signals. Classification from area allow type microplastic optical signals based on spectroscopy. In this paper, we investigate impact high-frequency noise performance related tasks. It is well-known that highly dependent peak visibility, but it also known signal smoothing a common step pre-processing measured This raises trade-off between preservation depends user-defined parameters. The results obtained work suggest linear discriminant analysis model cannot generalize properly presence noisy signals, whereas an error-correcting output codes better suited to account inherent noise. Moreover, principal components (PCA) can become must-do robust models, given its simplicity natural capabilities. Our study noise, possible use PCA as reduction technique addition dimensionality functionality are fundamental aspects work.

Language: Английский

---

Modeling the Temporal Evolution of Plastic Film Microplastics in Soil using a Backpropagation Neural Network

Journal of Hazardous Materials, Journal Year: 2024, Volume and Issue: 480, P. 136312 - 136312

Published: Oct. 30, 2024

Language: Английский

---

Multi-Perspective Interpretation for One-Dimensional Conventional Neural Network Model to Identify Iron-Bearing Waste Material

Published: Jan. 1, 2024

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Language: Английский

---