MYCOLab Ai Logo - gold and army green

Advancing automated identification of airborne fungal spores: guidelines for cultivation and reference dataset creation

Summary

Scientists developed systematic methods to grow and collect fungal spores in controlled conditions, then test them with automated air monitoring devices. Using two different monitoring systems that analyze spore images and fluorescence properties, they trained computer algorithms to recognize different fungal species. This work creates standardized guidelines that will help hospitals, allergy clinics, and agricultural services automatically detect and identify airborne fungal spores, which are important for managing allergies and plant diseases.

Background

Airborne fungal spores pose significant concerns for human and plant health, requiring precise monitoring systems. While European standards exist for manual volumetric monitoring, there is growing interest in automated real-time methods. However, these automated approaches rely heavily on machine learning and face challenges due to limited training data availability for fungal spores.

Objective

To outline best practices for cultivating fungal reference material, harvesting clean spores, aerosolizing them, and creating tailored datasets for training automated classification algorithms for airborne fungal spore identification.

Results

The SwisensPoleno Jupiter achieved macro average F1-scores of 0.77-0.83 with holographic and fluorescence data, while accuracies on Plair Rapid-E+ ranged from 83.4% to 95.1% with macro average F1-score of 0.61. Cladosporium and Curvularia species showed strong recognition, while some Alternaria species were more challenging to differentiate at the species level.

Conclusion

This protocol establishes standardized guidelines for fungal spore production and dataset creation that can facilitate more efficient, accurate, and automatic identification of airborne fungal spores, supporting operational bioaerosol monitoring networks.
  • Published in:Aerobiologia (Bologna),
  • Study Type:Experimental/Protocol Development,
  • Source: PMC12176942, PMID: 40547190, DOI: 10.1007/s10453-025-09864-y
Scroll to Top