A multi-step machine learning approach to assess the impact of COVID-19 lockdown on NO2 attributable deaths in Milan and Rome, Italy

Background Air pollution is one of the main concerns for the health of European citizens, and cities are currently striving to accomplish EU air pollution regulation. The 2020 COVID-19 lockdown measures can be seen as an unintended but effective experiment to assess the impact of traffic restriction policies on air pollution. Our objective was to estimate the impact of the lockdown measures on NO2 concentrations and health in the two largest Italian cities. Methods NO2 concentration datasets were built using data deriving from a 1-month citizen science monitoring campaign that took place in Milan and Rome just before the Italian lockdown period. Annual mean NO2 concentrations were estimated for a lockdown scenario (Scenario 1) and a scenario without lockdown (Scenario 2), by applying city-specific annual adjustment factors to the 1-month data. The latter were estimated deriving data from Air Quality Network stations and by applying a machine learning approach. NO2 spatial distribution was estimated at a neighbourhood scale by applying Land Use Random Forest models for the two scenarios. Finally, the impact of lockdown on health was estimated by subtracting attributable deaths for Scenario 1 and those for Scenario 2, both estimated by applying literature-based dose–response function on the counterfactual concentrations of 10 μg/m3. Results The Land Use Random Forest models were able to capture 41–42% of the total NO2 variability. Passing from Scenario 2 (annual NO2 without lockdown) to Scenario 1 (annual NO2 with lockdown), the population-weighted exposure to NO2 for Milan and Rome decreased by 15.1% and 15.3% on an annual basis. Considering the 10 μg/m3 counterfactual, prevented deaths were respectively 213 and 604. Conclusions Our results show that the lockdown had a beneficial impact on air quality and human health. However, compliance with the current EU legal limit is not enough to avoid a high number of NO2 attributable deaths. This contribution reaffirms the potentiality of the citizen science approach and calls for more ambitious traffic calming policies and a re-evaluation of the legal annual limit value for NO2 for the protection of human health.

Machine Learning Prediction of Progression in FEV1 in the COPDGene Study

Background: The heterogeneous nature of COPD complicates the identification of the predictors of disease progression and consequently the development of effective therapies. We aimed to improve the prediction of disease progression in COPD by using machine learning and incorporating a rich dataset of phenotypic features. Methods: We included 4,496 smokers with available data from their enrollment and 5-year follow-up visits in the Genetic Epidemiology of COPD (COPDGene) study. We constructed supervised random forest models to predict 5-year progression in FEV1 from 46 baseline demographic, clinical, physiologic, and imaging features. Using cross-validation, we randomly partitioned participants into training and testing samples. We also validated the results in the COPDGene 10-year follow-up visit. Results: Predicting the change in FEV1 over time is more challenging than simply predicting the future absolute FEV1 level. Nevertheless, the area under the ROC curves for the prediction of subjects in the top quartile of observed disease progression was 0.70 in the 10-year follow-up data. The model performance accuracy was best for GOLD1-2 subjects and it was harder to achieve accurate prediction in advanced stages of the disease. Predictive variables differed in their relative importance as well as for the predictions by GOLD grade. Conclusion: This state-of-the art approach along with deep phenotyping predicts FEV1 progression with reasonable accuracy. There is significant room for improvement in future models. This prediction model facilitates the identification of smokers at increased risk for rapid disease progression. Such findings may be useful in the selection of patient populations for targeted clinical trials.

Diagnostic and predictive accuracy of anti-mullerian hormone for ovarian function after chemotherapy in premenopausal women with early breast cancer

Purpose Accurate diagnosis and prediction of loss of ovarian function after chemotherapy for premenopausal women with early breast cancer (eBC) is important for future fertility and clinical decisions regarding the need for subsequent adjuvant ovarian suppression. We have investigated the value of anti-mullerian hormone (AMH) as serum biomarker for this. Methods AMH was measured in serial blood samples from 206 premenopausal women aged 40–45 years with eBC, before and at intervals after chemotherapy. The diagnostic accuracy of AMH for loss of ovarian function at 30 months after chemotherapy and the predictive value for that of AMH measurement at 6 months were analysed. Results Undetectable AMH showed a high diagnostic accuracy for absent ovarian function at 30 months with AUROC 0.89 (96% CI 0.84–0.94, P < 0.0001). PPV of undetectable AMH at 6 months for a menopausal estradiol level at 30 months was 0.77. In multivariate analysis age, pre-treatment AMH and FSH, and taxane treatment were significant predictors, and combined with AMH at 6 months, gave AUROC of 0.90 (95% CI 0.86–0.94), with PPV 0.79 for loss of ovarian function at 30 months. Validation by random forest models with 30% data retained gave similar results. Conclusions AMH is a reliable diagnostic test for lack of ovarian function after chemotherapy in women aged 40–45 with eBC. Early analysis of AMH after chemotherapy allows identification of women who will not recover ovarian function with good accuracy. These analyses will help inform treatment decisions regarding adjuvant endocrine therapy in women who were premenopausal before starting chemotherapy.

Machine learning applied to lentic habitat use by spawning walleye demonstrates the benefits of considering multiple spatial scales in aquatic research

Scale and hierarchy have received less attention in aquatic systems compared to terrestrial. Walleye Sander vitreus spawning habitat offers an opportunity to investigate scale’s importance. We estimated lake-, transect-, and quadrat-scale influences on nearshore walleye egg deposition in 28 Minnesota lakes from 2016-2018. Random forest models (RFM) estimated importance of predictive variables to walleye egg deposition. Predictive accuracies of a multi-scale classification tree (CT) and a quadrat-scale CT were compared. RFM results suggested that five of our variables were unimportant when predicting egg deposition. The multi-scale CT was more accurate than the quadrat-scale CT when predicting egg deposition. Both model results suggest that in-lake egg deposition by walleye is regulated by hierarchical abiotic processes and that silt/clay abundance at the transect-scale (reef-scale) is more important than abundance at the quadrat-scale (within-reef). Our results show machine learning can be used for scale-optimization and potentially to determine cross-scale interactions. Further incorporation of scale and hierarchy into studies of aquatic systems will increase our understanding of species-habitat relationships, especially in lentic systems where multi-scale approaches are rarely used.

Multimodal User Feedback During Adaptive Robot-Human Presentations

Feedback is an essential part of all communication, and agents communicating with humans must be able to both give and receive feedback in order to ensure mutual understanding. In this paper, we analyse multimodal feedback given by humans towards a robot that is presenting a piece of art in a shared environment, similar to a museum setting. The data analysed contains both video and audio recordings of 28 participants, and the data has been richly annotated both in terms of multimodal cues (speech, gaze, head gestures, facial expressions, and body pose), as well as the polarity of any feedback (negative, positive, or neutral). We train statistical and machine learning models on the dataset, and find that random forest models and multinomial regression models perform well on predicting the polarity of the participants' reactions. An analysis of the different modalities shows that most information is found in the participants' speech and head gestures, while much less information is found in their facial expressions, body pose and gaze. An analysis of the timing of the feedback shows that most feedback is given when the robot makes pauses (and thereby invites feedback), but that the more exact timing of the feedback does not affect its meaning.

Random Forest for Genomic Prediction

We give a detailed description of random forest and exemplify its use with data from plant breeding and genomic selection. The motivations for using random forest in genomic-enabled prediction are explained. Then we describe the process of building decision trees, which are a key component for building random forest models. We give (1) the random forest algorithm, (2) the main hyperparameters that need to be tuned, and (3) different splitting rules that are key for implementing random forest models for continuous, binary, categorical, and count response variables. In addition, many examples are provided for training random forest models with different types of response variables with plant breeding data. The random forest algorithm for multivariate outcomes is provided and its most popular splitting rules are also explained. In this case, some examples are provided for illustrating its implementation even with mixed outcomes (continuous, binary, and categorical). Final comments about the pros and cons of random forest are provided.

Spatial Distribution and Population Estimation of Dogs in Thailand: Implications for Rabies Prevention and Control

Poor management of dog populations causes many problems in different countries, including rabies. To strategically design a dog population management, certain sets of data are required, such as the population size and spatial distribution of dogs. However, these data are rarely available or incomplete. Hence, this study aimed to describe the characteristics of dog populations in Thailand, explore their spatial distribution and relevant factors, and estimate the number of dogs in the whole country. First, four districts were selected as representatives of each region. Each district was partitioned into grids with a 300-m resolution. The selected grids were then surveyed, and the number of dogs and related data were collected. Random forest models with a two-part approach were used to quantify the association between the surveyed dog population and predictor variables. The spatial distribution of dog populations was then predicted. A total of 1,750 grids were surveyed (945 grids with dog presence and 805 grids with dog absence). Among the surveyed dogs, 86.6% (12,027/13,895) were owned. Of these, 51% were classified as independent, followed by confined (25%), semi-independent (21%), and unidentified dogs (3%). Seventy-two percent (1,348/1,868) of the ownerless dogs were feral, and the rest were community dogs. The spatial pattern of the dog populations was highly distributed in big cities such as Bangkok and its suburbs. In owned dogs, it was linked to household demographics, whereas it was related to community factors in ownerless dogs. The number of estimated dogs in the entire country was 12.8 million heads including 11.2 million owned dogs (21.7 heads/km2) and 1.6 million ownerless dogs (3.2 heads/km2). The methods developed here are extrapolatable to a larger area and use much less budget and manpower compared to the present practices. Our results are helpful for canine rabies prevention and control programs, such as dog population management and control and rabies vaccine allocation.

Classifying Oscillatory Brain Activity Associated With Indian Rasas Using Network Metrics

Neural oscillations are the rich source to understand cognition, perception, and emotions. Decades of research on brain oscillations have primarily discussed neural signatures for the western classification of emotions. Despite this, the Indian ancient treatise on emotions popularly known as Rasas has remained unexplored. In this study, we collected Electroencephalography (EEG) encodings while participants watched nine emotional movie clips corresponding to nine Rasas. The key objective of this study is to identify the brain waves that could distinguish between Rasas. Therefore, we decompose the EEG signals into five primary frequency bands comprising delta (1-4 Hz), theta (4-7 Hz), alpha (8-13 Hz), beta (13-30 Hz), and gamma (30-45 Hz). We construct the functional networks from EEG time-series data and subsequently utilize the fourteen graph-theoretical measures to compute the features. Random Forest models are trained on the extracted features, and we present our findings based on classifier predictions. We observe slow (delta) and fast brain waves (beta and gamma) exhibited the maximum discriminating features between Rasas, whereas alpha and theta bands showed fewer distinguishable pairs. Out of nine Rasas, Sringaram, Bibhatsam, and Bhayanakam displayed the most distinguishing characteristics from other Rasas. Interestingly, our results are consistent with the previous studies, which highlight the significant role of higher frequency oscillations for the classification of emotions. Our finding on the alpha band is consistent with the previous study, which reports the maximum similarity in brain networks across emotions in the alpha band. This research contributes to the pioneering work on Indian Rasas utilizing brain responses.