@article{10902/36217,
year = {2025},
month = {9},
url = {https://hdl.handle.net/10902/36217},
abstract = {Cyanobacterial blooms pose significant environmental and public health risks due to the production of toxins that contaminate water sources and disrupt aquatic ecosystems. Rapid and accurate identification of cyanobacterial species is crucial for effective monitoring and management strategies. In this study, we combined Raman spectroscopy with deep learning techniques to classify four toxic cyanobacterial species: Dolichospermum crassum, Aphanizomenon sp., Planktothrix agardhii and Microcystis aeruginosa. Spectral data were acquired using a confocal Raman microscope with a 532 nm excitation wavelength and subjected to preprocessing and filtering to enhance signal quality. We evaluated a multichannel one-dimensional convolutional neural network (1D-CNN) approach that incorporates raw spectra, baseline estimations, and preprocessed spectra. This multichannel approach improved overall classification accuracy, achieving 86% compared to 74% with a traditional single-channel 1D-CNN using only preprocessed spectra while maintaining low overfitting. Shapley Additive exPlanations (SHAP) were applied to identify critical spectral regions for classification to enhance interpretability. These findings highlight the potential of combining Raman spectroscopy with explainable deep learning methods as a powerful tool for water quality monitoring and the early detection of Harmful Algal Blooms (HABs).},
organization = {This work was supported by the R+D projects PREVAL23/05, INNVAL23/10, and INNVAL24/28, funded by Instituto de Investigación Marqués de Valdecilla (IDIVAL) ; J.F.A. acknowledges RYC2022-035279-I, funded by MCIN/AEI/10.13039/501100011033 and FSE+; TED2021-130378B-C21, funded by MCIN/AEI/10.13039/501100011033 and European Union NextGenerationEU/PRTR; PID2022-137269OB-C22, fun ded by MCIN/AEI/10.13039/501100011033 and FEDER, UE; Plan Nacional de I+D+i and Instituto de Salud Carlos III (ISCIII), Subdirección General de Redes Centros de Investigación Cooperativa, Ministerio de Ciencia, Innovación Universidades, through CIBER-BBN (CB16/01/00430) and CIBERINFEC (CB21/13/00068), co-financed by the European Regional Development Fund “A way to achieve Europe”.
We thank the Biology Department of Universidad Autónoma de Madrid for their support and for providing the samples used in this study.},
publisher = {Elsevier},
publisher = {Talanta, 2025, 292, 127845},
title = {Comprehensive Raman spectroscopy analysis for differentiating toxic cyanobacteria through multichannel 1D-CNNs and SHAP-based explainability},
author = {Fernández Manteca, María Gabriela and García García, Borja and Deus Álvarez, Susana and Goméz Galdós, Celia and Pérez Asensio, Andrea and Algorri Genaro, José Francisco and Monteoliva Herreras, Agustín and López Higuera, José Miguel and Rodríguez Cobo, Luis and Ocampo Sosa, Alain Antonio and Cobo García, Adolfo},
}