Using Baidu index to nowcast hand-foot-mouth disease in China: a meta learning approach

Using Baidu Index Data to Improve Chickenpox Surveillance in Yunnan, China: Infodemiology Study

BACKGROUND

Chickenpox is an old but easily neglected infectious disease. Although chickenpox is preventable by vaccines, vaccine breakthroughs often occur, and the chickenpox epidemic is on the rise. Chickenpox is not included in the list of regulated communicable diseases that must be reported and controlled by public and health departments; therefore, it is crucial to rapidly identify and report varicella outbreaks during the early stages. The Baidu index (BDI) can supplement the traditional surveillance system for infectious diseases, such as brucellosis and dengue, in China. The number of reported chickenpox cases and internet search data also showed a similar trend. BDI can be a useful tool to display the outbreak of infectious diseases.

OBJECTIVE

This study aimed to develop an efficient disease surveillance method that uses BDI to assist in traditional surveillance.

METHODS

Chickenpox incidence data (weekly from January 2017 to June 2021) reported by the Yunnan Province Center for Disease Control and Prevention were obtained to evaluate the relationship between the incidence of chickenpox and BDI. We applied a support vector machine regression (SVR) model and a multiple regression prediction model with BDI to predict the incidence of chickenpox. In addition, we used the SVR model to predict the number of chickenpox cases from June 2021 to the first week of April 2022.

RESULTS

The analysis showed that there was a close correlation between the weekly number of newly diagnosed cases and the BDI. In the search terms we collected, the highest Spearman correlation coefficient was 0.747. Most BDI search terms, such as "chickenpox," "chickenpox treatment," "treatment of chickenpox," "chickenpox symptoms," and "chickenpox virus," trend consistently. Some BDI search terms, such as "chickenpox pictures," "symptoms of chickenpox," "chickenpox vaccine," and "is chickenpox vaccine necessary," appeared earlier than the trend of "chickenpox virus." The 2 models were compared, the SVR model performed better in all the applied measurements: fitting effect, R²=0.9108, root mean square error (RMSE)=96.2995, and mean absolute error (MAE)=73.3988; and prediction effect, R²=0.548, RMSE=189.1807, and MAE=147.5412. In addition, we applied the SVR model to predict the number of reported cases weekly in Yunnan from June 2021 to April 2022 using the same period of the BDI. The results showed that the fluctuation of the time series from July 2021 to April 2022 was similar to that of the last year and a half with no change in the level of prevention and control.

CONCLUSIONS

These findings indicated that the BDI in Yunnan Province can predict the incidence of chickenpox in the same period. Thus, the BDI is a useful tool for monitoring the chickenpox epidemic and for complementing traditional monitoring systems.

Forecasting HFMD Cases Using Weather Variables and Google Search Queries in Sabah, Malaysia

HFMD is a viral-mediated infectious illness of increasing public health importance. This study aimed to develop a forecasting tool utilizing climatic predictors and internet search queries for informing preventive strategies in Sabah, Malaysia. HFMD case data from the Sabah State Health Department, climatic predictors from the Malaysia Meteorological Department, and Google search trends from the Google trends platform between the years 2010-2018 were utilized. Cross-correlations were estimated in building a seasonal auto-regressive moving average (SARIMA) model with external regressors, directed by measuring the model fit. The selected variables were then validated using test data utilizing validation metrics such as the mean average percentage error (MAPE). Google search trends evinced moderate positive correlations to the HFMD cases (r_0-6weeks: 0.47-0.56), with temperature revealing weaker positive correlations (r_0-3weeks: 0.17-0.22), with the association being most intense at 0-1 weeks. The SARIMA model, with regressors of mean temperature at lag 0 and Google search trends at lag 1, was the best-performing model. It provided the most stable predictions across the four-week period and produced the most accurate predictions two weeks in advance (RMSE = 18.77, MAPE = 0.242). Trajectorial forecasting oscillations of the model are stable up to four weeks in advance, with accuracy being the highest two weeks prior, suggesting its possible usefulness in outbreak preparedness.

Explanation of hand, foot, and mouth disease cases in Japan using Google Trends before and during the COVID-19: infodemiology study

Background

Coronavirus Disease 2019 (COVID-19) pandemic affects common diseases, but its impact on hand, foot, and mouth disease (HFMD) is unclear. Google Trends data is beneficial for approximate real-time statistics and because of ease in access, is expected to be used for infection explanation from an information-seeking behavior perspective. We aimed to explain HFMD cases before and during COVID-19 using Google Trends.

Methods

HFMD cases were obtained from the National Institute of Infectious Diseases, and Google search data from 2009 to 2021 in Japan were downloaded from Google Trends. Pearson correlation coefficients were calculated between HFMD cases and the search topic “HFMD” from 2009 to 2021. Japanese tweets containing “HFMD” were retrieved to select search terms for further analysis. Search terms with counts larger than 1000 and belonging to ranges of infection sources, susceptible sites, susceptible populations, symptoms, treatment, preventive measures, and identified diseases were retained. Cross-correlation analyses were conducted to detect lag changes between HFMD cases and search terms before and during the COVID-19 pandemic. Multiple linear regressions with backward elimination processing were used to identify the most significant terms for HFMD explanation.

Results

HFMD cases and Google search volume peaked around July in most years, excluding 2020 and 2021. The search topic “HFMD” presented strong correlations with HFMD cases, except in 2020 when the COVID-19 outbreak occurred. In addition, the differences in lags for 73 (72.3%) search terms were negative, which might indicate increasing public awareness of HFMD infections during the COVID-19 pandemic. The results of multiple linear regression demonstrated that significant search terms contained the same meanings but expanded informative search content during the COVID-19 pandemic.

Conclusions

The significant terms for the explanation of HFMD cases before and during COVID-19 were different. Awareness of HFMD infections in Japan may have improved during the COVID-19 pandemic. Continuous monitoring is important to promote public health and prevent resurgence. The public interest reflected in information-seeking behavior can be helpful for public health surveillance.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-022-07790-9.

An improved strabismus screening method with combination of meta-learning and image processing under data scarcity

Purpose

Considering the scarcity of normal and strabismic images, this study proposed a method that combines a meta-learning approach with image processing methods to improve the classification accuracy when meta-learning alone is used for screening strabismus.

Methods

The meta-learning approach was first pre-trained on a public dataset to obtain a well-generalized embedding network to extract distinctive features of images. On the other hand, the image processing methods were used to extract the position features of eye regions (e.g., iris position, corneal light reflex) as supplementary features to the distinctive features. Afterward, principal component analysis was applied to reduce the dimensionality of distinctive features for integration with low-dimensional supplementary features. The integrated features were then used to train a support vector machine classifier for performing strabismus screening. Sixty images (30 normal and 30 strabismus) were used to verify the effectiveness of the proposed method, and its classification performance was assessed by computing the accuracy, specificity, and sensitivity through 5,000 experiments.

Results

The proposed method achieved a classification accuracy of 0.805 with a sensitivity (correct classification of strabismus) of 0.768 and a specificity (correct classification of normal) of 0.842, whereas the classification accuracy of using meta-learning alone was 0.709 with a sensitivity of 0.740 and a specificity of 0.678.

Conclusion

The proposed strabismus screening method achieved promising classification accuracy and gained significant accuracy improvement over using meta-learning alone under data scarcity.

Improving Google Flu Trends for COVID-19 estimates using Weibo posts

While incomplete non-medical data has been integrated into prediction models for epidemics, the accuracy and the generalizability of the data are difficult to guarantee. To comprehensively evaluate the ability and applicability of using social media data to predict the development of COVID-19, a new confirmed case prediction algorithm improving the Google Flu Trends algorithm is established, called Weibo COVID-19 Trends (WCT), based on the post dataset generated by all users in Wuhan on Sina Weibo. A genetic algorithm is designed to select the keyword set for filtering COVID-19 related posts. WCT can constantly outperform the highest average test score in the training set between daily new confirmed case counts and the prediction results. It remains to produce the best prediction results among other algorithms when the number of forecast days increases from one to eight days with the highest correlation score from 0.98 (P ​< 0.01) to 0.86 (P ​< 0.01) during all analysis period. Additionally, WCT effectively improves the Google Flu Trends algorithm's shortcoming of overestimating the epidemic peak value. This study offers a highly adaptive approach for feature engineering of third-party data in epidemic prediction, providing useful insights for the prediction of newly emerging infectious diseases at an early stage.

Forecasting Teleconsultation Demand Using an Ensemble CNN Attention-Based BILSTM Model with Additional Variables

To enhance the forecasting accuracy of daily teleconsultation demand, this study proposes an ensemble hybrid deep learning model. The proposed ensemble CNN attention-based BILSTM model (ECA-BILSTM) combines shallow convolutional neural networks (CNNs), attention mechanisms, and bidirectional long short-term memory (BILSTM). Moreover, additional variables are selected according to the characteristics of teleconsultation demand and added to the inputs of forecasting models. To verify the superiority of ECA-BILSTM and the effectiveness of additional variables, two actual teleconsultation datasets collected in the National Telemedicine Center of China (NTCC) are used as the experimental data. Results showed that ECA-BILSTMs can significantly outperform corresponding benchmark models. And two key additional variables were identified for teleconsultation demand prediction improvement. Overall, the proposed ECA-BILSTM model with effective additional variables is a feasible promising approach in teleconsultation demand forecasting.

Predicting seasonal influenza using supermarket retail records

Increased availability of epidemiological data, novel digital data streams, and the rise of powerful machine learning approaches have generated a surge of research activity on real-time epidemic forecast systems. In this paper, we propose the use of a novel data source, namely retail market data to improve seasonal influenza forecasting. Specifically, we consider supermarket retail data as a proxy signal for influenza, through the identification of sentinel baskets, i.e., products bought together by a population of selected customers. We develop a nowcasting and forecasting framework that provides estimates for influenza incidence in Italy up to 4 weeks ahead. We make use of the Support Vector Regression (SVR) model to produce the predictions of seasonal flu incidence. Our predictions outperform both a baseline autoregressive model and a second baseline based on product purchases. The results show quantitatively the value of incorporating retail market data in forecasting models, acting as a proxy that can be used for the real-time analysis of epidemics.

Using deep learning to predict the hand-foot-and-mouth disease of enterovirus A71 subtype in Beijing from 2011 to 2018

Hand-foot-and-month disease (HFMD), especially the enterovirus A71 (EV-A71) subtype, is a major health problem in Beijing, China. Previous studies mainly used regressive models to forecast the prevalence of HFMD, ignoring its intrinsic age groups. This study aims to predict HFMD of EV-A71 subtype in three age groups (0–3, 3–6 and > 6 years old) from 2011 to 2018 using residual-convolutional-recurrent neural network (CNNRNN-Res), convolutional-recurrent neural network (CNNRNN) and recurrent neural network (RNN). They were compared with auto-regressio, global auto-regression and vector auto-regression on both short-term and long-term prediction. Results showed that CNNRNN-Res and RNN had higher accuracies on point forecast tasks, as well as robust performances in long-term prediction. Three deep learning models also had better skills in peak intensity forecast, and CNNRNN-Res achieved the best results in the peak month forecast. We also found that three age groups had consistent outbreak trends and similar patterns of prediction errors. These results highlight the superior performance of deep learning models in HFMD prediction and can assist the decision-makers to refine the HFMD control measures according to age groups.

Search trends and prediction of human brucellosis using Baidu index data from 2011 to 2018 in China

Reporting on brucellosis, a relatively rare infectious disease caused by Brucella, is often delayed or incomplete in traditional disease surveillance systems in China. Internet search engine data related to brucellosis can provide an economical and efficient complement to a conventional surveillance system because people tend to seek brucellosis-related health information from Baidu, the largest search engine in China. In this study, brucellosis incidence data reported by the CDC of China and Baidu index data were gathered to evaluate the relationship between them. We applied an autoregressive integrated moving average (ARIMA) model and an ARIMA model with Baidu search index data as the external variable (ARIMAX) to predict the incidence of brucellosis. The two models based on brucellosis incidence data were then compared, and the ARIMAX model performed better in all the measurements we applied. Our results illustrate that Baidu index data can enhance the traditional surveillance system to monitor and predict brucellosis epidemics in China.

Forecast of the trend in incidence of acute hemorrhagic conjunctivitis in China from 2011-2019 using the Seasonal Autoregressive Integrated Moving Average (SARIMA) and Exponential Smoothing (ETS) models

BACKGROUND

This study aimed to explore the demographic and distributive features of acute hemorrhagic conjunctivitis (AHC). We constructed seasonal autoregressive integrated moving average (SARIMA) and exponential smoothing (ETS) models to predict its trend in incidence in mainland China and provided evidence for the government to formulate policies regarding AHC prevention.

METHODS

Data regarding the distribution and demographic characteristics of AHC in China from 2011-2016 were retrieved from the Public Health Scientific Data website. Monthly AHC data from January 2011 to June 2019 were used to establish and evaluate the SARIMA and ETS models.

RESULTS

During 2011-2016, a total of 213,325 cases were reported; 46.01% were farmers, patients aged ≤9 years had the highest risk, and the male:female ratio was 1.31:1. Guangxi and Guangdong had the highest number of reported AHC cases. The SARIMA (0, 0, 1) (2, 0, 0) ₁₂ model with the minimum root mean squared error and mean absolute percentage error were finally selected for in-sample simulation.

CONCLUSIONS

AHC remains a serious public health problem in Southern and Eastern China that mainly affects farmers and children younger than 9 years. It is recommended that the health administration strengthen the publicity and education regarding AHC prevention among farmers and accelerate the development of related vaccines and treatment measures.

Study of the Effects of Air Pollutants on Human Health Based on Baidu Indices of Disease Symptoms and Air Quality Monitoring Data in Beijing, China

There is an increasing body of evidence showing the impact of air pollutants on human health such as on the respiratory, and cardio- and cerebrovascular systems. In China, as people begin to pay more attention to air quality, recent research focused on the quantitative assessment of the effects of air pollutants on human health. To assess the health effects of air pollutants and to construct an indicator placing emphasis on health impact, a generalized additive model was selected to assess the health burden caused by air pollution. We obtained Baidu indices (an evaluation indicator launched by Baidu Corporation to reflect the search popularity of keywords from its search engine) to assess daily query frequencies of 25 keywords considered associated with air pollution-related diseases. Moreover, we also calculated the daily concentrations of major air pollutants (including PM10, PM2.5, SO₂, O₃, NO₂, and CO) and the daily air quality index (AQI) values, and three meteorological factors: daily mean wind level, daily mean air temperature, and daily mean relative humidity. These data cover the area of Beijing from 1 March 2015 to 30 April 2017. Through the analysis, we produced the relative risks (RRs) of the six main air pollutants for respiratory, and cardio- and cerebrovascular diseases. The results showed that O₃ and NO₂ have the highest health impact, followed by PM10 and PM2.5. The effects of any pollutant on cardiovascular diseases was consistently higher than on respiratory diseases. Furthermore, we evaluated the currently used AQI in China and proposed an RR-based index (health AQI, HAQI) that is intended for better indicating the effects of air pollutants on respiratory, and cardio- and cerebrovascular diseases than AQI. A higher Pearson correlation coefficient between HAQI and RRTotal than that between AQI and RRTotal endorsed our efforts.