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

Abstract

Airborne bioparticles, including fungal spores, are of major concern for human and plant health, necessitating precise monitoring systems. While a European norm exists for manual volumetric monitoring, there's a growing interest in automated real-time methods. However, these methods rely heavily on machine learning, facing challenges due to diverse particle characteristics and limited training data availability, especially for fungal spores. This study aims to address this gap by outlining best practices for collecting reference material and creating tailored datasets for training algorithms. Using 17 fungal species from the Belgian fungi collection BCCM/IHEM, including five Alternaria species, key aspects such as in vitro cultivation, dry spore harvest, and aerosolization were addressed. Simple classification models were developed, achieving varying accuracies on different monitors. The Plair Rapid-E+ demonstrated accuracies ranging from 83.4% to 95.1% (macro average F1-score 0.61), with better recognition for Cladosporium spp. and Curvularia caricae-papayae. The SwisensPoleno Jupiter, initially achieving a macro average F1-score of 0.77 with holographic images of eight genera, improved to 0.83 when combined with fluorescence data. Accuracies ranged from 55 to 95%, with notable performance for Alternaria spp. and Curvularia caricae-papayae. Species differentiation was also shown to be possible for Cladosporium, but was more difficult for some Alternaria species, while the macro average F1-score remained good (0.72). Overall, this protocol paves the way for more efficient, standard, and accurate automatic identification of airborne fungal spores.

Supplementary information: The online version contains supplementary material available at 10.1007/s10453-025-09864-y.

Keywords: Airflow cytometry; Automatic identification; Culture collection; Fungal spores; Machine learning.

Fig. 1

A Instructions for constructing a plastic cyclone collector vial using a micro-tube and two micropipette tips, B Picture of an assembled cyclone collector vial, C Picture of serially connected three cyclone collector vials for increasing the harvesting efficiency of large particles, D Scheme of the harvesting system to aspirate dry spores from fungi culture on Petri dish

Fig. 2

Experimental laboratory setup for the aerosolization of fungal spores and their detection by the SwisensPoleno Jupiter. Homogenous aerosolization within the chamber of the SwisensAtomizer can be performed either by using an open square cuvette or the cyclone collector vial, filled with dry spore material. Arrows indicate the direction of air flows to and out of the container

Fig. 3

Holography images (120 × 120 µm) of representative single spore and datasets of average relative fluorescence (stacked barplots) recorded by the SwisensPoleno Jupiter, for 17 selected fungal species. Non-scaled representative drawings of spores adapted from Simmons EG, 2007 and the Atlas of Clinical Fungi (https://www.atlasclinicalfungi.org). Data are shown prior to any filtering for the creation of training datasets

Fig. 4

Experimental laboratory setup for the aerosolization of fungal spores and their detection by the Plair Rapid-E+. Arrows indicate the direction of air flows to and out of the cyclone collector vial

Fig. 5

Confusion matrices of the recognition performance of the convolutional neural networks trained on (A) holography imaging data or (B) combined holography imaging and fluorescence data from 17 fungal spore taxa recorded by the SwisensPoleno Jupiter. Percentage of accurately predicted and misclassified test particles is presented as a heat map

Fig. 6

Confusion matrix of the recognition performance of the convolutional neural networks trained on fluorescence data from seven fungal spore genera recorded by the Plair Rapid-E+. Alternaria spp. include the mixed species A. alternata, A. arborescens, A. botrytis, A. chartarum and A. terricola. Cladosporium spp. include the mixed species Cladosporium cladosporioides, Cladosporium herbarum and Cladosporium sphaerospermum. Percentage of accurately predicted and misclassified test particles is presented as a heat map