Toward the use of neural networks for influenza prediction at multiple spatial resolutions

Commuting-driven competition between transmission chains shapes seasonal influenza virus epidemics in the United States

Despite intensive study, much remains unknown about the dynamics of seasonal influenza virus epidemic establishment and spread in the United States (US) each season. By reconstructing transmission lineages from seasonal influenza virus genomes collected in the US from 2014 to 2023, we show that most epidemics consisted of multiple distinct transmission lineages. Spread of these lineages exhibited strong spatiotemporal hierarchies and lineage size was correlated with timing of lineage establishment in the US. Mechanistic epidemic simulations suggest that mobility-driven competition between lineages determined the extent of individual lineages' geographical spread. Based on phylogeographic analyses and epidemic simulations, lineage-specific movement patterns were dominated by human commuting behavior. These results suggest that given the locations of early-season epidemic sparks, the topology of inter-state human mobility yields repeatable patterns of which influenza viruses will circulate where, but the importance of short-term processes limits predictability of regional and national epidemics.

Teaser

Epidemics consist of multiple sub-epidemics that compete for susceptible hosts and spread due to the movement of commuters.

Spatial analysis and risk mapping of Fasciola hepatica infection in dairy cattle at the Peruvian central highlands

This study aimed to develop maps for Fasciola hepatica infection occurrence in dairy cattle in the districts of Matahuasi and Baños in the Peruvian central highlands. For this, a model based on the correlation between environmental variables and the prevalence of infection was constructed. Flukefinder® coprological test were performed in samples from dairy cattle from 8 herds, during both the rainy and wet season. Grazing plots were geo-referenced to obtain information on environmental variables. Monthly temperature, monthly rainfall, elevation, slope, normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), normalized difference water index (NDWI), distance to rivers, urban areas and roads were obtained by using remote sensor images and ArcGIS®. Multilayer perceptron Artificial Neural Networks modeling were applied to construct a predictive model for the occurrence of fasciolosis, based on the relationship between environmental variables and level of infection. Kappa coefficient (k > 0.6) was used to evaluate concordance between observed and forecasted risk by the model. Coprological results demonstrated an average prevalence from 20% to 100%, in Matahuasi, and between 0 and 87.5%, in Baños. A model with a high level of concordance between predicted and observed infection risk (k = 0.77) was obtained, having as major predicting variables: slope, NDWI, NDVI and EVI. Fasciolosis risk was categorized as low (p < 20%), medium (20% < p < 50%) and high (p ≥ 50%) level. Using ArcGIS 10.4.1, risk maps were developed for each risk level of fasciolosis. Maps of fasciolosis occurrence showed that 87.2% of Matahuasi area presented a high risk for bovine fasciolosis during the dry season, and 76.6% in the wet season. In contrast, 21.9% of Baños area had a high risk of infection during the dry season and 12.1% during the wet season. In conclusion, our model showed areas with high risk for fasciolosis occurrence in both districts during both dry and rainy periods. Slope, NDWI, NDVI and EVI were the major predictors for fasciolosis occurrence.

Neural network models for influenza forecasting with associated uncertainty using Web search activity trends

Influenza affects millions of people every year. It causes a considerable amount of medical visits and hospitalisations as well as hundreds of thousands of deaths. Forecasting influenza prevalence with good accuracy can significantly help public health agencies to timely react to seasonal or novel strain epidemics. Although significant progress has been made, influenza forecasting remains a challenging modelling task. In this paper, we propose a methodological framework that improves over the state-of-the-art forecasting accuracy of influenza-like illness (ILI) rates in the United States. We achieve this by using Web search activity time series in conjunction with historical ILI rates as observations for training neural network (NN) architectures. The proposed models incorporate Bayesian layers to produce associated uncertainty intervals to their forecast estimates, positioning themselves as legitimate complementary solutions to more conventional approaches. The best performing NN, referred to as the iterative recurrent neural network (IRNN) architecture, reduces mean absolute error by 10.3% and improves skill by 17.1% on average in nowcasting and forecasting tasks across 4 consecutive flu seasons.

Psychological AI: Designing Algorithms Informed by Human Psychology

Psychological artificial intelligence (AI) applies insights from psychology to design computer algorithms. Its core domain is decision-making under uncertainty, that is, ill-defined situations that can change in unexpected ways rather than well-defined, stable problems, such as chess and Go. Psychological theories about heuristic processes under uncertainty can provide possible insights. I provide two illustrations. The first shows how recency-the human tendency to rely on the most recent information and ignore base rates-can be built into a simple algorithm that predicts the flu substantially better than did Google Flu Trends's big-data algorithms. The second uses a result from memory research-the paradoxical effect that making numbers less precise increases recall-in the design of algorithms that predict recidivism. These case studies provide an existence proof that psychological AI can help design efficient and transparent algorithms.

Graph Neural Network Modeling of Web Search Activity for Real-time Pandemic Forecasting

The utilization of web search activity for pandemic forecasting has significant implications for managing disease spread and informing policy decisions. However, web search records tend to be noisy and influenced by geographical location, making it difficult to develop large-scale models. While regularized linear models have been effective in predicting the spread of respiratory illnesses like COVID-19, they are limited to specific locations. The lack of incorporation of neighboring areas' data and the inability to transfer models to new locations with limited data has impeded further progress. To address these limitations, this study proposes a novel self-supervised message-passing neural network (SMPNN) framework for modeling local and cross-location dynamics in pandemic forecasting. The SMPNN framework utilizes an MPNN module to learn cross-location dependencies through self-supervised learning and improve local predictions with graph-generated features. The framework is designed as an end-to-end solution and is compared with state-of-the-art statistical and deep learning models using COVID-19 data from England and the US. The results of the study demonstrate that the SMPNN model outperforms other models by achieving up to a 6.9% improvement in prediction accuracy and lower prediction errors during the early stages of disease outbreaks. This approach represents a significant advancement in disease surveillance and forecasting, providing a novel methodology, datasets, and insights that combine web search data and spatial information. The proposed SMPNN framework offers a promising avenue for modeling the spread of pandemics, leveraging both local and cross-location information, and has the potential to inform public health policy decisions.

Predicting influenza with pandemic-awareness via Dynamic Virtual Graph Significance Networks

Every year, influenza spreads worldwide and burdens people's health substantially. We need a reliable model to help hospitals, pharmaceutical companies, and governments better prepare for influenza outbreaks in a timely manner. However, the domain knowledge for such public health events, such as the variable influenza seasonality and occasional pandemics, poses significant challenges in predicting influenza outbreaks. The existing methods use current and historical values in a user-defined time window as input to predict future values but lack considering the situations outside the window. To address these limitations, we proposed Dynamic Virtual Graph Significance Networks (DVGSN). The graph-based algorithm can supervisedly and dynamically learn the implied knowledge from similar "infection situations" in all the historical timepoints without the limitation of time window. Furthermore, representation learning on the dynamic virtual graph can tackle the varied seasonality with pandemic-awareness without requiring domain knowledge input. The extensive experiments on real-world influenza data demonstrate that DVGSN significantly outperforms the state-of-the-art methods. To the best of our knowledge, this is the first attempt to supervisedly learn a dynamic virtual graph for time-series prediction tasks. Moreover, the proposed method has rich interpretabilities, which makes the method more acceptable in the fields of public health, life sciences, and so on. Our source code and dataset are available at https://github.com/aI-area/DVGSN.

Influenza's Plummeting During the COVID-19 Pandemic: The Roles of Mask-Wearing, Mobility Change, and SARS-CoV-2 Interference

Seasonal influenza activity typically peaks in the winter months but plummeted globally during the current coronavirus disease 2019 (COVID-19) pandemic. Unraveling lessons from influenza's unprecedented low profile is critical in informing preparedness for incoming influenza seasons. Here, we explored a country-specific inference model to estimate the effects of mask-wearing, mobility changes (international and domestic), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) interference in China, England, and the United States. We found that a one-week increase in mask-wearing intervention had a percent reduction of 11.3%-35.2% in influenza activity in these areas. The one-week mobility mitigation had smaller effects for the international (1.7%-6.5%) and the domestic community (1.6%-2.8%). In 2020-2021, the mask-wearing intervention alone could decline percent positivity by 13.3-19.8. The mobility change alone could reduce percent positivity by 5.2-14.0, of which 79.8%-98.2% were attributed to the deflected international travel. Only in 2019-2020, SARS-CoV-2 interference had statistically significant effects. There was a reduction in percent positivity of 7.6 (2.4-14.4) and 10.2 (7.2-13.6) in northern China and England, respectively. Our results have implications for understanding how influenza evolves under non-pharmaceutical interventions and other respiratory diseases and will inform health policy and the design of tailored public health measures.

Influenza's Plummeting During the COVID-19 Pandemic: The Roles of Mask-Wearing, Mobility Change, and SARS-CoV-2 Interference

Seasonal influenza activity typically peaks in the winter months but plummeted globally during the current coronavirus disease 2019 (COVID-19) pandemic. Unraveling lessons from influenza's unprecedented low profile is critical in informing preparedness for incoming influenza seasons. Here, we explored a country-specific inference model to estimate the effects of mask-wearing, mobility changes (international and domestic), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) interference in China, England, and the United States. We found that a one-week increase in mask-wearing intervention had a percent reduction of 11.3%-35.2% in influenza activity in these areas. The one-week mobility mitigation had smaller effects for the international (1.7%-6.5%) and the domestic community (1.6%-2.8%). In 2020-2021, the mask-wearing intervention alone could decline percent positivity by 13.3-19.8. The mobility change alone could reduce percent positivity by 5.2-14.0, of which 79.8%-98.2% were attributed to the deflected international travel. Only in 2019-2020, SARS-CoV-2 interference had statistically significant effects. There was a reduction in percent positivity of 7.6 (2.4-14.4) and 10.2 (7.2-13.6) in northern China and England, respectively. Our results have implications for understanding how influenza evolves under non-pharmaceutical interventions and other respiratory diseases and will inform health policy and the design of tailored public health measures.

Using digital traces to build prospective and real-time county-level early warning systems to anticipate COVID-19 outbreaks in the United States

Coronavirus disease 2019 (COVID-19) continues to affect the world, and the design of strategies to curb disease outbreaks requires close monitoring of their trajectories. We present machine learning methods that leverage internet-based digital traces to anticipate sharp increases in COVID-19 activity in U.S. counties. In a complementary direction to the efforts led by the Centers for Disease Control and Prevention (CDC), our models are designed to detect the time when an uptrend in COVID-19 activity will occur. Motivated by the need for finer spatial resolution epidemiological insights, we build upon previous efforts conceived at the state level. Our methods-tested in an out-of-sample manner, as events were unfolding, in 97 counties representative of multiple population sizes across the United States-frequently anticipated increases in COVID-19 activity 1 to 6 weeks before local outbreaks, defined when the effective reproduction number R_t becomes larger than 1 for a period of 2 weeks.

Mapping of machine learning approaches for description, prediction, and causal inference in the social and health sciences

Machine learning (ML) methodology used in the social and health sciences needs to fit the intended research purposes of description, prediction, or causal inference. This paper provides a comprehensive, systematic meta-mapping of research questions in the social and health sciences to appropriate ML approaches by incorporating the necessary requirements to statistical analysis in these disciplines. We map the established classification into description, prediction, counterfactual prediction, and causal structural learning to common research goals, such as estimating prevalence of adverse social or health outcomes, predicting the risk of an event, and identifying risk factors or causes of adverse outcomes, and explain common ML performance metrics. Such mapping may help to fully exploit the benefits of ML while considering domain-specific aspects relevant to the social and health sciences and hopefully contribute to the acceleration of the uptake of ML applications to advance both basic and applied social and health sciences research.

Forecasting Weekly Dengue Cases by Integrating Google Earth Engine-Based Risk Predictor Generation and Google Colab-Based Deep Learning Modeling in Fortaleza and the Federal District, Brazil

Efficient and accurate dengue risk prediction is an important basis for dengue prevention and control, which faces challenges, such as downloading and processing multi-source data to generate risk predictors and consuming significant time and computational resources to train and validate models locally. In this context, this study proposed a framework for dengue risk prediction by integrating big geospatial data cloud computing based on Google Earth Engine (GEE) platform and artificial intelligence modeling on the Google Colab platform. It enables defining the epidemiological calendar, delineating the predominant area of dengue transmission in cities, generating the data of risk predictors, and defining multi-date ahead prediction scenarios. We implemented the experiments based on weekly dengue cases during 2013-2020 in the Federal District and Fortaleza, Brazil to evaluate the performance of the proposed framework. Four predictors were considered, including total rainfall (R_sum), mean temperature (T_mean), mean relative humidity (RH_mean), and mean normalized difference vegetation index (NDVI_mean). Three models (i.e., random forest (RF), long-short term memory (LSTM), and LSTM with attention mechanism (LSTM-ATT)), and two modeling scenarios (i.e., modeling with or without dengue cases) were set to implement 1- to 4-week ahead predictions. A total of 24 models were built, and the results showed in general that LSTM and LSTM-ATT models outperformed RF models; modeling could benefit from using historical dengue cases as one of the predictors, and it makes the predicted curve fluctuation more stable compared with that only using climate and environmental factors; attention mechanism could further improve the performance of LSTM models. This study provides implications for future dengue risk prediction in terms of the effectiveness of GEE-based big geospatial data processing for risk predictor generation and Google Colab-based risk modeling and presents the benefits of using historical dengue data as one of the input features and the attention mechanism for LSTM modeling.

Non-von Neumann multi-input spike signal processing enabled by an artificial synaptic multiplexer

Multiplexing is essential for technologies that require processing of a large amount of information in real time. Here, we present an artificial synaptic multiplexing unit capable of realizing parallel multi-input control system. Ion gel was used as a dielectric layer of the artificial synaptic multiplexing unit because of its ionic property, allowing multigating for parallel input. A closed-loop control system that enables multi-input-based feedback for actuator bending control was realized by incorporating an ion gel-based artificial synaptic multiplexing unit, an actuator, and a bending angle sensor. The proposed multi-input control system could simultaneously process input and feedback signals, offering a breakthrough in industries in which the processing of vast amounts of streaming data is essential.

Predicting Nanoparticle Delivery to Tumors Using Machine Learning and Artificial Intelligence Approaches

Background

Low delivery efficiency of nanoparticles (NPs) to the tumor is a critical barrier in the field of cancer nanomedicine. Strategies on how to improve NP tumor delivery efficiency remain to be determined.

Methods

This study analyzed the roles of NP physicochemical properties, tumor models, and cancer types in NP tumor delivery efficiency using multiple machine learning and artificial intelligence methods, using data from a recently published Nano-Tumor Database that contains 376 datasets generated from a physiologically based pharmacokinetic (PBPK) model.

Results

The deep neural network model adequately predicted the delivery efficiency of different NPs to different tumors and it outperformed all other machine learning methods; including random forest, support vector machine, linear regression, and bagged model methods. The adjusted determination coefficients (R2) in the full training dataset were 0.92, 0.77, 0.77 and 0.76 for the maximum delivery efficiency (DEmax), delivery efficiency at 24 h (DE24), at 168 h (DE168), and at the last sampling time (DETlast). The corresponding R2 values in the test dataset were 0.70, 0.46, 0.33 and 0.63, respectively. Also, this study showed that cancer type was an important determinant for the deep neural network model in predicting the tumor delivery efficiency across all endpoints (19-29%). Among all physicochemical properties, the Zeta potential and core material played a greater role than other properties, such as the type, shape, and targeting strategy.

Conclusion

This study provides a quantitative model to improve the design of cancer nanomedicine with greater tumor delivery efficiency. These results help to improve our understanding of the causes of low NP tumor delivery efficiency. This study demonstrates the feasibility of integrating artificial intelligence with PBPK modeling approaches to study cancer nanomedicine.

The Incoming Influenza Season - China, the United Kingdom, and the United States, 2021-2022

INTRODUCTION

Seasonal influenza activity has declined globally since the widespread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission. There has been scarce information to understand the future dynamics of influenza - and under different hypothesis on relaxation of non-pharmaceutical interventions (NPIs) in particular - after the disruptions to seasonal patterns.

METHODS

We collected data from public sources in China, the United Kingdom, and the United States, and forecasted the influenza dynamics in the incoming 2021-2022 season under different NPIs. We considered Northern China and Southern China separately, due to the sharp difference in the patterns of seasonal influenza. For the United Kingdom, data were collected for England only.

RESULTS

Compared to the epidemics in 2017-2019, longer and blunter influenza outbreaks could occur should NPIs be fully lifted, with percent positivity varying from 10.5 to 18.6 in the studying regions. The rebounds would be smaller if the mask-wearing intervention continued or the international mobility stayed low, but sharper if the mask-wearing intervention was lifted in the middle of influenza season. Further, influenza activity could stay low under a much less stringent mask-wearing intervention coordinated with influenza vaccination.

CONCLUSIONS

The results added to our understandings of future influenza dynamics after the global decline during the coronavirus disease 2019 (COVID-19) pandemic. In light of the uncertainty on the incoming circulation strains and the relatively low negative impacts of mask wearing on society, our findings suggested that wearing mask could be considered as an accompanying mitigation measure in influenza prevention and control, especially for seasons after long periods of low-exposure to influenza viruses.