This readme file was generated on [2025-11-13] by [Irene Teixido-Orries] GENERAL INFORMATION Title of Dataset: Visible-near infrared spectroscopy and near-infrared hyperspectral imaging for the detection of T-2 and HT-2 toxins in individual oat grains Author Name: Irene Teixido-Orries ORCID: 0000-0002-7405-6272 Institution & Theme: Applied Mycology Unit, University of Lleida-Agrotecnio Address: Av. Rovira Roure 191, 25198, Lleida, Spain Email: irene.teixido@udl.cat Author/Associate or Co-investigator Information Name: Francisco Molino ORCID: 0000-0002-1286-3482 Institution & Theme: Applied Mycology Unit, University of Lleida-Agrotecnio Address: Av. Rovira Roure 191, 25198, Lleida, Spain Email: francisco.molino@udl.cat Author//Principal Investigator/Supervisor Information Name: Carol Verheecke-Vaessen ORCID: 0000-0003-3470-9803 Institution & Theme: Magan Centre of Applied Mycology, Cranfield University, Address: Cranfield, MK43 0AL, United Kingdom Email: c.verheecke@cranfield.ac.uk Date of data collection: 2023-12-01 2024-05-01 Geographic location of data collection: Cranfield, MK43 0AL, United Kingdom Lleida, 25198, Spain Information about funding sources that supported the collection of the data: This work was supported by the Spanish Ministry of Science and Innovation (predoctoral grant FPU21/00073 and Project PID2020- 114836RB-I00 funded by MCIN/AEI/10.13039/501100011033) and Cranfield University. SHARING/ACCESS INFORMATION Licenses/restrictions placed on the data: None Links to publications that cite or use the data: https://doi.org/10.1016/j.foodcont.2025.111676 Links to other publicly accessible locations of the data: N/A Links/relationships to ancillary data sets: No Was data derived from another source? If yes, list source(s): No Recommended citation for this dataset: Teixido-Orries, I., Molino, F., Agusti-Fernandez, P. (2025). Visible-near infrared spectroscopy and near-infrared hyperspectral imaging for the detection of T-2 and HT-2 toxins in individual oat grains [Dataset]. DATA & FILE OVERVIEW File List: 1. HT-2+T-2.csv Chemical quantification of T-2 and HT-2 toxins in individual oat grains using LC-MS/MS (200 grains). 2. VIS-NIRSpectra.csv Visible–NIR reflectance spectra (351–2500 nm) of 600 individual oat grain scans. 3. NIR-HSISpectra.csv NIR-HSI mean spectra (893.1–1731 nm) extracted from NIR-hyperspectral imaging (600 individual oat grain scans). Relationship between files, if important: All three files correspond to the same experimental set of oat grains. HT-2+T-2.csv provides reference chemical values. VIS-NIRSpectra.csv provides Vis-NIR point-based spectral features. NIR-HSISpectra.csv provides imaging-based spectral features. Together, they support model development for predicting T-2/HT-2 contamination. Additional related data collected that was not included in the current data package: Raw HSI datacubes and original instrument files (not included here due to size). Are there multiple versions of the dataset? No If yes, name of file(s) that was updated: - Why was the file updated? - When was the file updated? - METHODOLOGICAL INFORMATION Description of methods used for collection/generation of data: Vis-NIR spectroscopy: Measurements performed on individual oat grains using a Vis–NIR spectrometer (351–2500 nm). NIR-HSI imaging: Hyperspectral images acquired between 893.1–1731 nm. Mean spectra extracted for each grain. Chemical analysis (T-2+HT-2): LC-MS/MS quantification after extraction and clean-up following validated laboratory protocols. Methodology details can be found in https://doi.org/10.1016/j.foodcont.2025.111676. Methods for processing the data: VIS-NIR and NIR-HSI spectral ranges were trimmed to remove noisy extremes. Raw, SNV-corrected, derivative and combined preprocessing strategies were applied. LC-MS/MS data were processed with Analyst® 1.6.3 and MultiQuant™ 3.0.3. All spectral datasets were aligned by grain ID and sample ID. Toxin concentrations were expressed in ng/g (single-grain basis). Data were organised into three final datasets: 1. Chemical quantification (200 grains) 2. Vis-NIR spectra (600 spectra) 3. NIR-HSI spectra (600 spectra) Instrument- or software-specific information needed to interpret the data: VIS-NIR: LabSpec 2500®, Indico Pro software HSI: Resonon Pika NIR-320, Spectronon PRO LC-MS/MS: Exion LC + Sciex 6500+, Analyst® 1.6.3, MultiQuant™ 3.0.3 Data processing: Python 3.11.10 (NumPy, SciPy, Pandas, Scikit-Learn, Imbalanced-Learn), RStudio 4.2.3, Microsoft Excel 16.87 Standards and calibration information, if appropriate: T-2 and HT-2 standards (100 µg/mL) prepared in acetonitrile. Nine-level matrix-matched calibration curves analysed in duplicate. Calibration acceptance criterion: R² ≥ 0.98. Instrument calibrations: VIS-NIR and HSI: dark and white references LC-MS/MS: ammonium adduct–forming solvent system Environmental/experimental conditions: HSI measurements performed after 20 min lamp warm-up. VIS-NIR system stabilised for 50 min before scanning. Temperature maintained at ~25°C during extraction and centrifugation. Describe any quality-assurance procedures performed on the data: Blank matrix, matrix-matched standards and spiked samples analysed for selectivity. Recoveries and precision tested across three validation days. LOD and LOQ determined using the pseudo-blank approach. Grain segmentation for HSI manually checked to ensure accuracy. Baseline and white calibrations performed regularly during spectral acquisition. People involved with sample collection, processing, analysis and/or submission: Irene Teixido-Orries: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Francisco Molino: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Data curation, Conceptualization. Pau Agusti-Fernandez: Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation. Ebenezer Ayibiowu: Methodology, Investigation, Conceptualization. Derek Croucher: Writing – review & editing, Methodology. Angel Medina: Supervision, Project administration, Funding acquisition. Sonia Marín: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Funding acquisition. Carol Verheecke-Vaessen: Writing – review & editing, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization. DATA-SPECIFIC INFORMATION FOR: HT-2+T-2.csv Number of variables: 8 Number of cases/rows: 201 Variable List: - Grain number: codification of the grain number - Sample: codification of the samples - T-2 toxin in the grain (ng/g): concentration of T-2 toxin in the oat grains - HT-2 toxin in the grain (ng/g): concentration of HT-2 toxin in the oat grains - Sum T-2+HT-2 (ng/g): sum of the concentration of T-2 and HT-2 toxins in the oat grains - U/L 1250 ppb: grains above (U) or below (L) the legal limit for T-2 and HT-2 toxins (1250 ppb) - U/L 1000 ppb: grains above (U) or below (L) the previous indicative level for T-2 and HT-2 toxins (1000 ppb) - U/L >10000 ppb: grains above (U) or below (L) a high content of T-2 and HT-2 toxins (10000 ppb) Missing data codes: