Integration of reference data from different Rapid-E devices supports automatic pollen detection in more locations

Size-classified monitoring of ATP bioluminescence for rapid assessment of biological distribution in airborne particulates

The COVID-19 pandemic ignited massive research into the rapid detection of bioaerosols. In particular, nanotechnology-based detection strategies are proposed as alternatives because of issues in bioaerosol enrichment and lead time for molecular diagnostics; however, the practical implementation of such techniques is still unclear due to obstacles regarding the large research and development effort and investment for the validation. The use of adenosine triphosphate (ATP) bioluminescence (expressed as relative luminescence unit (RLU) per unit volume of air) of airborne particulate matter (PM) to determine the bacterial population as a representative of the total bioaerosols (viruses, bacteria, and fungi) has been raised frequently because of the high reponse speed, resolution, and compatibility with culture-based bioaerosol monitoring. On the other hand, additional engineering attempts are required to confer significance because of the size-classified (bioluminescence for different PM sizes) and specific (bioluminescence per unit PM mass) biological risks of air for providing proper interventions in the case of airborne transmission. In this study, disc-type impactors to cut-off aerosols larger than 1 μm, 2.5 μm, and 10 μm were designed and constructed to collect PM1, PM2.5, and PM10 on sampling swabs. This engineering enabled reliable size-classified bioluminescence signals using a commercial ATP luminometer after just 5 min of air intake. The simultaneous operations of a six-stage Andersen impactor and optical PM spectrometers were conducted to determine the correlations between the resulting RLU and colony forming unit (CFU; from the Andersen impactor) or PM mass concentration (deriving specific bioluminescence).

Explainable AI for unveiling deep learning pollen classification model based on fusion of scattered light patterns and fluorescence spectroscopy

Pollen monitoring have become data-intensive in recent years as real-time detectors are deployed to classify airborne pollen grains. Machine learning models with a focus on deep learning, have an essential role in the pollen classification task. Within this study we developed an explainable framework to unveil a deep learning model for pollen classification. Model works on data coming from single particle detector (Rapid-E) that records for each particle optical fingerprint with scattered light and laser induced fluorescence. Morphological properties of a particle are sensed with the light scattering process, while chemical properties are encoded with fluorescence spectrum and fluorescence lifetime induced by high-resolution laser. By utilizing these three data modalities, scattering, spectrum, and lifetime, deep learning-based models with millions of parameters are learned to distinguish different pollen classes, but a proper understanding of such a black-box model decisions demands additional methods to employ. Our study provides the first results of applied explainable artificial intelligence (xAI) methodology on the pollen classification model. Extracted knowledge on the important features that attribute to the predicting particular pollen classes is further examined from the perspective of domain knowledge and compared to available reference data on pollen sizes, shape, and laboratory spectrofluorometer measurements.