Pulse Brain · Growing Health Evidence Index
Tier 3 — Observational / field trialPeer-reviewedSupporting research

Near-infrared hyperspectral imaging as a rapid tool for detection of toxin-producing microalgae in seawater

Catarina Moreirinha; Sergey Kucheryavskiy; Maria João Botelho; M Salvande Fraga; André Sobrinho-Gonçalves; Bárbara Frazão; Alisa M. Rudnitskaya

Marine Environmental Research · 2026

Read source ↗ All evidence

Summary

This study demonstrates that near-infrared hyperspectral imaging (NIR-HSI) can rapidly and non-destructively detect toxin-producing microalgae directly in natural seawater at ecologically relevant abundances. Using weekly to bi-weekly samples from a Portuguese lagoon collected over one year, multivariate image analysis models distinguished toxin-producing groups with cross-validated accuracies of 0.87–0.98, and outperformed FT-MIR spectroscopy for several toxin classes. The findings suggest NIR-HSI has strong potential as a monitoring technology for harmful algal bloom programmes.

Regional applicability

This technological approach is directly applicable to United Kingdom coastal and estuarine monitoring, particularly for shellfish safety and HAB surveillance programmes administered by Cefas and local authorities. The method's rapid turnaround and non-destructive nature could enhance early-warning capacity in UK coastal waters, though validation with local phytoplankton communities and toxin profiles would strengthen implementation.

Key measures

Classification accuracy (cross-validated), toxin-producer detection limits, cell counts by Utermöhl microscopy reference method, spectral data in 900–1700 nm range, PLS-DA model performance

Outcomes reported

The study evaluated NIR-HSI as a rapid screening tool for detecting five groups of toxin-producing microalgae (AST, PST, DST, YTX, AZT producers) in natural seawater samples collected over one year from a Portuguese lagoon. Cross-validated classification models achieved 0.87–0.98 accuracy in distinguishing samples above and below toxin detection thresholds, with species-level discrimination within Dinophysis reaching 0.98 accuracy.

Supporting research

Exact primary PubMed abstract describes near-infrared/machine-learning monitoring of toxin-producing microalgae in Ria de Aveiro. Official IPMA documentation ties monitoring these harmful species and marine biotoxins to shellfish harvesting safety; this is a documented external food-safety application.

Limits: One-year estuarine monitoring/method study; food link is supported by the official monitoring context, not a measured consumer dietary outcome. It does not establish seafood toxin concentration, human exposure, regulatory fitness or generalisation.

Theme
Measurement & metrics
Subject
Measurement methods & nutrient profiling
Study type
Research
Study design
Field trial with multivariate image analysis and supervised classification
Source type
Peer-reviewed study
Status
Published
Geography
Portugal
System type
Aquaculture
DOI
10.1016/j.marenvres.2026.108373
Catalogue ID
NRmusgqbdk-03x

Topic tags

Pulse AI · ask about this record

Dig deeper with Pulse AI.

Ask about this record, its theme or its relevance to UK farming and policy. Pulse AI uses selected catalogue evidence and cites the sources it draws on.