Airborne pollen grain detection from partially labelled data utilising semi-supervised learning.

Airborne pollen monitoring has been conducted for more than a century now, as knowledge of the quantity and periodicity of airborne pollen has diverse use cases, like reconstructing historic climates and tracking current climate change, forensic applications, and up to warning those affected by pollen-induced respiratory allergies. Hence, related work on automation of pollen classification already exists. In contrast, detection of pollen is still conducted manually, and it is the gold standard for accuracy. So, here we used a new-generation, automated, near-real-time pollen monitoring sampler, the BAA500, and we used data consisting of both raw and synthesised microscope images. Apart from the automatically generated, commercially-labelled data of all pollen taxa, we additionally used manual corrections to the pollen taxa, as well as a manually created test set of bounding boxes and pollen taxa, so as to more accurately evaluate the real-life performance. For the pollen detection, we employed two-stage deep neural network object detectors. We explored a semi-supervised training scheme to remedy the partial labelling. Using a teacher-student approach, the model can add pseudo-labels to complete the labelling during training. To evaluate the performance of our deep learning algorithms and to compare them to the commercial algorithm of the BAA500, we created a manual test set, in which an expert aerobiologist corrected automatically annotated labels. For the novel manual test set, both the supervised and semi-supervised approaches clearly outperform the commercial algorithm with an F1 score of up to 76.9 % compared to 61.3 %. On an automatically created and partially labelled test dataset, we obtain a maximum mAP of 92.7 %. Additional experiments on raw microscope images show comparable performance for the best models, which potentially justifies reducing the complexity of the image generation process. Our results bring automatic pollen monitoring a step forward, as they close the gap in pollen detection performance between manual and automated procedure.

Detecting Airborne Pollen Using an Automatic, Real-Time Monitoring System: Evidence from Two Sites.

Airborne pollen monitoring has been an arduous task, making ecological applications and allergy management virtually disconnected from everyday practice. Over the last decade, intensive research has been conducted worldwide to automate this task and to obtain real-time measurements. The aim of this study was to evaluate such an automated biomonitoring system vs. the conventional 'gold-standard' Hirst-type technique, attempting to assess which may more accurately provide the genuine exposure to airborne pollen. Airborne pollen was monitored in Augsburg since 2015 with two different methods, a novel automatic Bio-Aerosol Analyser, and with the conventional 7-day recording Hirst-type volumetric trap, in two different sites. The reliability, performance, accuracy, and comparability of the BAA500 Pollen Monitor (PoMo) vs. the conventional device were investigated, by use of approximately 2.5 million particles sampled during the study period. The observations made by the automated PoMo showed an average accuracy of approximately 85%. However, it also exhibited reliability problems, with information gaps within the main pollen season of between 17 to 19 days. The PoMo automated algorithm had identification issues, mainly confusing the taxa of Populus, Salix and Tilia. Hirst-type measurements consistently exhibited lower pollen abundances (median of annual pollen integral: 2080), however, seasonal traits were more comparable, with the PoMo pollen season starting slightly later (median: 3 days), peaking later (median: 5 days) but also ending later (median: 14 days). Daily pollen concentrations reported by Hirst-type traps vs. PoMo were significantly, but not closely, correlated (r = 0.53-0.55), even after manual classification. Automatic pollen monitoring has already shown signs of efficiency and accuracy, despite its young age; here it is suggested that automatic pollen monitoring systems may be more effective in capturing a larger proportion of the airborne pollen diversity. Even though reliability issues still exist, we expect that this new generation of automated bioaerosol monitoring will eventually change the aerobiological era, as known for almost 70 years now.

Climate change and global health: A call to more research and more action.

There is increasing understanding, globally, that climate change and increased pollution will have a profound and mostly harmful effect on human health. This review brings together international experts to describe both the direct (such as heat waves) and indirect (such as vector-borne disease incidence) health impacts of climate change. These impacts vary depending on vulnerability (i.e., existing diseases) and the international, economic, political, and environmental context. This unique review also expands on these issues to address a third category of potential longer-term impacts on global health: famine, population dislocation, and environmental justice and education. This scholarly resource explores these issues fully, linking them to global health in urban and rural settings in developed and developing countries. The review finishes with a practical discussion of action that health professionals around the world in our field can yet take.

Atmospheric pollutants and their association with olive and grass aeroallergen concentrations in Córdoba (Spain).

Cumulative data indicate that pollen grains and air pollution reciprocally interact. Climate changes seem also to influence pollen allergenicity. Depending on the plant species and on the pollutant type and concentration, this interaction may modify the features and metabolism of the pollen grain. Previous results revealed a significant positive correlation between pollen and aeroallergen, even using two different samplers. However, some discrepancy days have been also detected with low pollen but high aeroallergen concentrations. The main aim of the present paper is to find how the environmental factors, and specially pollutants, could affect the amount of allergens from olive and grass airborne pollen. Pollen grains were collected by a Hirst-type volumetric spore trap. Aeroallergen was simultaneously sampled by a low-volume Cyclone Burkard sampler. Phl p 5 and Ole e 1 aeroallergen were quantified by double-sandwich ELISA test. The data related to air pollutants, pollen grains, and aeroallergens were analyzed with descriptive statistic. Spearman's correlation test was used to identify potential correlations between these variables. There is a significant positive correlation between aeroallergens and airborne pollen concentrations, in both studied pollen types, so allergen concentrations could be explained with the pollen concentration. The days with unlinked events coincide between olive and grass allergens. Nevertheless, concerning to our results, pollutants do not affect the amount of allergens per pollen. Even if diverse pollutants show an unclear relationship with the allergen concentration, this association seems to be a casual effect of the leading role of some meteorological parameters.