Relationship between Exhaled Aerosol and Carbon Dioxide Emission Across Respiratory Activities.

Respiratory particles produced during vocalized and nonvocalized activities such as breathing, speaking, and singing serve as a major route for respiratory pathogen transmission. This work reports concomitant measurements of exhaled carbon dioxide volume (VCO2) and minute ventilation (VE), along with exhaled respiratory particles during breathing, exercising, speaking, and singing. Exhaled CO2 and VE measured across healthy adult participants follow a similar trend to particle number concentration during the nonvocalized exercise activities (breathing at rest, vigorous exercise, and very vigorous exercise). Exhaled CO2 is strongly correlated with mean particle number (r = 0.81) and mass (r = 0.84) emission rates for the nonvocalized exercise activities. However, exhaled CO2 is poorly correlated with mean particle number (r = 0.34) and mass (r = 0.12) emission rates during activities requiring vocalization. These results demonstrate that in most real-world environments vocalization loudness is the main factor controlling respiratory particle emission and exhaled CO2 is a poor surrogate measure for estimating particle emission during vocalization. Although measurements of indoor CO2 concentrations provide valuable information about room ventilation, such measurements are poor indicators of respiratory particle concentrations and may significantly underestimate respiratory particle concentrations and disease transmission risk.

A comparison of respiratory particle emission rates at rest and while speaking or exercising.

Background: The coronavirus disease-19 (COVID-19) pandemic led to the prohibition of group-based exercise and the cancellation of sporting events. Evaluation of respiratory aerosol emissions is necessary to quantify exercise-related transmission risk and inform mitigation strategies. Methods: Aerosol mass emission rates are calculated from concurrent aerosol and ventilation data, enabling absolute comparison. An aerodynamic particle sizer (0.54-20 µm diameter) samples exhalate from within a cardiopulmonary exercise testing mask, at rest, while speaking and during cycle ergometer-based exercise. Exercise challenge testing is performed to replicate typical gym-based exercise and very vigorous exercise, as determined by a preceding maximally exhaustive exercise test. Results: We present data from 25 healthy participants (13 males, 12 females; 36.4 years). The size of aerosol particles generated at rest and during exercise is similar (unimodal ~0.57-0.71 µm), whereas vocalization also generated aerosol particles of larger size (i.e. was bimodal ~0.69 and ~1.74 µm). The aerosol mass emission rate during speaking (0.092 ng s-1; minute ventilation (VE) 15.1 L min-1) and vigorous exercise (0.207 ng s-1, p = 0.726; VE 62.6 L min-1) is similar, but lower than during very vigorous exercise (0.682 ng s-1, p < 0.001; VE 113.6 L min-1). Conclusions: Vocalisation drives greater aerosol mass emission rates, compared to breathing at rest. Aerosol mass emission rates in exercise rise with intensity. Aerosol mass emission rates during vigorous exercise are no different from speaking at a conversational level. Mitigation strategies for airborne pathogens for non-exercise-based social interactions incorporating vocalisation, may be suitable for the majority of exercise settings. However, the use of facemasks when exercising may be less effective, given the smaller size of particles produced.

Comparing aerosol number and mass exhalation rates from children and adults during breathing, speaking and singing.

Aerosol particles of respirable size are exhaled when individuals breathe, speak and sing and can transmit respiratory pathogens between infected and susceptible individuals. The COVID-19 pandemic has brought into focus the need to improve the quantification of the particle number and mass exhalation rates as one route to provide estimates of viral shedding and the potential risk of transmission of viruses. Most previous studies have reported the number and mass concentrations of aerosol particles in an exhaled plume. We provide a robust assessment of the absolute particle number and mass exhalation rates from measurements of minute ventilation using a non-invasive Vyntus Hans Rudolf mask kit with straps housing a rotating vane spirometer along with measurements of the exhaled particle number concentrations and size distributions. Specifically, we report comparisons of the number and mass exhalation rates for children (12-14 years old) and adults (19-72 years old) when breathing, speaking and singing, which indicate that child and adult cohorts generate similar amounts of aerosol when performing the same activity. Mass exhalation rates are typically 0.002-0.02 ng s-1 from breathing, 0.07-0.2 ng s-1 from speaking (at 70-80 dBA) and 0.1-0.7 ng s-1 from singing (at 70-80 dBA). The aerosol exhalation rate increases with increasing sound volume for both children and adults when both speaking and singing.

Pre-Synaptic Pool Modification (PSPM): A supervised learning procedure for recurrent spiking neural networks.

Learning synaptic weights of spiking neural network (SNN) models that can reproduce target spike trains from provided neural firing data is a central problem in computational neuroscience and spike-based computing. The discovery of the optimal weight values can be posed as a supervised learning task wherein the weights of the model network are chosen to maximize the similarity between the target spike trains and the model outputs. It is still largely unknown whether optimizing spike train similarity of highly recurrent SNNs produces weight matrices similar to those of the ground truth model. To this end, we propose flexible heuristic supervised learning rules, termed Pre-Synaptic Pool Modification (PSPM), that rely on stochastic weight updates in order to produce spikes within a short window of the desired times and eliminate spikes outside of this window. PSPM improves spike train similarity for all-to-all SNNs and makes no assumption about the post-synaptic potential of the neurons or the structure of the network since no gradients are required. We test whether optimizing for spike train similarity entails the discovery of accurate weights and explore the relative contributions of local and homeostatic weight updates. Although PSPM improves similarity between spike trains, the learned weights often differ from the weights of the ground truth model, implying that connectome inference from spike data may require additional constraints on connectivity statistics. We also find that spike train similarity is sensitive to local updates, but other measures of network activity such as avalanche distributions, can be learned through synaptic homeostasis.