Using an MSRD neural network model on hospital data to provide infectious-disease early warnings (Preprint)

2021 ◽  
Author(s):  
MengYing Wang ◽  
Cuixia Lee ◽  
Cheng Yang ◽  
Yingyun Yang
Keyword(s):  
Public Health ◽  
Neural Network ◽  
Infectious Disease ◽  
Infectious Diseases ◽  
Neural Network Model ◽  
Prediction Accuracy ◽  
Hospital Data ◽  
Seir Model ◽  
Network Methods ◽  
National Public Health

BACKGROUND This study focuses on analyzing real data from a hospital to provide timely warnings of known infectious diseases with a view to actively preventing epidemics. OBJECTIVE The aim is to design MSRD model to predict the epidemic trend of infectious diseases based on real hospital data. METHODS Based on the daily reported data of infectious diseases between 2012–2020 from a large Chinese hospital, we selected seven common infectious diseases and constructed a Multi Self-regression Deep (MSRD) neural network model. This model, which is based on a recurrent neural network, can effectively model the epidemic trend of infectious diseases while considering the current influential factors and characteristics of historical development when calculating time-series data. The mean absolute error (MAE) and the root mean square error (RMSE) are used to evaluate the model’s fit and prediction accuracy. RESULTS We compared the MSRD model proposed in this study with the infectious disease SEIR-model using the national public health dataset on COVID-19 and another in-hospital infectious disease, namely, Hand-Foot-and-Mouth disease (HFMD). In an experiment with the national public health dataset, the MSRD proposed in this study demonstrated better performance than the SEIR model, which is because of the SEIR model being limited by factors such as the latent population. The SEIR model is hard to apply to real-world hospital scenarios. Our MSRD model is compared with other neural network methods. The dataset is from real hospital medical records for January 2012–December 2020. The MAE of the MSRD neural network for HFMD and influenza was as low as 0.6928 and 1.3782, respectively. In addition, our MSRD model was compared against other neural network methods such as SVM, Lasso, and Bayes; the MAE and RMSE were both better than those of other neural networks. CONCLUSIONS Our MSRD neural network has high prediction accuracy and can predict the development trend of infectious diseases on a daily basis. The MSRD model can act as a hospital infectious-disease early-warning system.

scholarly journals Patterns of smallpox mortality in London, England, over three centuries

PLoS Biology ◽  
2020 ◽  
Vol 18 (12) ◽  
pp. e3000506
Author(s):  
Olga Krylova ◽  
David J. D. Earn
Keyword(s):  
Public Health ◽  
Infectious Disease ◽  
Time Series ◽  
Spectral Analysis ◽  
Infectious Diseases ◽  
Health Policies ◽  
Human Populations ◽  
Public Health Policies ◽  
History Of ◽  
History Of Control

Smallpox is unique among infectious diseases in the degree to which it devastated human populations, its long history of control interventions, and the fact that it has been successfully eradicated. Mortality from smallpox in London, England was carefully documented, weekly, for nearly 300 years, providing a rare and valuable source for the study of ecology and evolution of infectious disease. We describe and analyze smallpox mortality in London from 1664 to 1930. We digitized the weekly records published in the London Bills of Mortality (LBoM) and the Registrar General’s Weekly Returns (RGWRs). We annotated the resulting time series with a sequence of historical events that might have influenced smallpox dynamics in London. We present a spectral analysis that reveals how periodicities in reported smallpox mortality changed over decades and centuries; many of these changes in epidemic patterns are correlated with changes in control interventions and public health policies. We also examine how the seasonality of reported smallpox mortality changed from the 17th to 20th centuries in London.

Climate Change-Associated Conflict and Infectious Disease

2017 ◽  
pp. 68-88
Author(s):  
Devin C. Bowles
Keyword(s):  
Public Health ◽  
Climate Change ◽  
Infectious Disease ◽  
Infectious Diseases ◽  
Disease Surveillance ◽  
Health Personnel ◽  
Forced Migration ◽  
Scarce Resources ◽  
Food Scarcity ◽  
Surveillance Programs

One of the least appreciated mechanisms by which climate change will affect infectious diseases is via increased violent conflict. Climate change will diminish agricultural and pastoral resources and increase food scarcity in many areas, including already impoverished equatorial regions. Many in the defence and public health fields anticipate that climate change will increase conflict by fuelling competition over scarce resources. Already, some commentators argue that the conflicts in Darfur and Syria were partially caused or exacerbated by climate change. Conflict facilitates a range of conditions conducive to the spread of many infectious diseases, including malnutrition, forced migration, unhygienic living conditions and widespread sexual assault. Flight or killing of health personnel inhibits vaccination, vector control and disease surveillance programs. Emergence of new diseases may go undetected and discovery of outbreaks could be suppressed for strategic reasons. These conditions combine to increase the risk of pandemics.

scholarly journals Exploiting the Relationship between Pruning Ratio and Compression Effect for Neural Network Model Based on TensorFlow

2020 ◽  
Vol 2020 ◽  
pp. 1-8
Author(s):  
Bo Liu ◽  
Qilin Wu ◽  
Yiwen Zhang ◽  
Qian Cao
Keyword(s):  
Neural Network ◽  
Network Model ◽  
Neural Network Model ◽  
Prediction Accuracy ◽  
Computing Time ◽  
Experimental Results ◽  
Data Set ◽  
Compression Effect ◽  
Model Based ◽  
The Relationship

Pruning is a method of compressing the size of a neural network model, which affects the accuracy and computing time when the model makes a prediction. In this paper, the hypothesis that the pruning proportion is positively correlated with the compression scale of the model but not with the prediction accuracy and calculation time is put forward. For testing the hypothesis, a group of experiments are designed, and MNIST is used as the data set to train a neural network model based on TensorFlow. Based on this model, pruning experiments are carried out to investigate the relationship between pruning proportion and compression effect. For comparison, six different pruning proportions are set, and the experimental results confirm the above hypothesis.

scholarly journals A DBN-Based Deep Neural Network Model with Multitask Learning for Online Air Quality Prediction

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Jiangeng Li ◽  
Xingyang Shao ◽  
Rihui Sun
Keyword(s):  
Neural Network ◽  
Air Quality ◽  
Network Model ◽  
Neural Network Model ◽  
Prediction Accuracy ◽  
Deep Neural Network ◽  
Multitask Learning ◽  
Air Pollutant ◽  
Quality Prediction ◽  
Air Quality Prediction

To avoid the adverse effects of severe air pollution on human health, we need accurate real-time air quality prediction. In this paper, for the purpose of improve prediction accuracy of air pollutant concentration, a deep neural network model with multitask learning (MTL-DBN-DNN), pretrained by a deep belief network (DBN), is proposed for forecasting of nonlinear systems and tested on the forecast of air quality time series. MTL-DBN-DNN model can solve several related prediction tasks at the same time by using shared information contained in the training data of different tasks. In the model, DBN is used to learn feature representations. Each unit in the output layer is connected to only a subset of units in the last hidden layer of DBN. Such connection effectively avoids the problem that fully connected networks need to juggle the learning of each task while being trained, so that the trained networks cannot get optimal prediction accuracy for each task. The sliding window is used to take the recent data to dynamically adjust the parameters of the MTL-DBN-DNN model. The MTL-DBN-DNN model is evaluated with a dataset from Microsoft Research. Comparison with multiple baseline models shows that the proposed MTL-DBN-DNN achieve state-of-art performance on air pollutant concentration forecasting.

BP Neural Network Model for early Diagnosis of Kawasaki Disease

2012 ◽  
Vol 468-471 ◽  
pp. 723-726 ◽  
Author(s):  
Jiang Huang ◽  
Jian Feng Chen
Keyword(s):  
Neural Network ◽  
Kawasaki Disease ◽  
Early Diagnosis ◽  
Infectious Diseases ◽  
Network Model ◽  
Neural Network Model ◽  
Bp Neural Network ◽  
Clinical Symptoms ◽  
Laboratory Data ◽  
Bp Neural Network Model

In order to diagnose Kawasaki Disease during early phase, clinical symptoms (temperature, rash, conjunctival injection, erythema of thelips, and oral mucosal changes) and laboratory data (white blood cell, neutrophil, platelet, high sensitive c-reactive protein, and erythrocyte sedimentation rate) of 138 children with Kawasaki disease or infectious diseases were used to develop a BP neural network model. 90 random cases were trained using MATLAB software for setting up the BP neural network model. The other 48 cases were analyzed to predict Kawasaki disease using this model. Results showed that the predict accuracy in patients with Kawasaki disease and children with infectious diseases are 95.6% and 88%, respectively. Our result indicates that the BP neural network model is likely to provide an accurate test for early diagnosis of Kawasaki disease.

scholarly journals An Inventory Controlled Supply Chain Model Based on Improved BP Neural Network

2013 ◽  
Vol 2013 ◽  
pp. 1-7 ◽  
Author(s):  
Wei He
Keyword(s):  
Neural Network ◽  
Supply Chain ◽  
Convergence Rate ◽  
Inventory Control ◽  
Network Model ◽  
Neural Network Model ◽  
Bp Neural Network ◽  
Prediction Accuracy ◽  
Inventory Level ◽  
Bp Neural Network Model

Inventory control is a key factor for reducing supply chain cost and increasing customer satisfaction. However, prediction of inventory level is a challenging task for managers. As one of the widely used techniques for inventory control, standard BP neural network has such problems as low convergence rate and poor prediction accuracy. Aiming at these problems, a new fast convergent BP neural network model for predicting inventory level is developed in this paper. By adding an error offset, this paper deduces the new chain propagation rule and the new weight formula. This paper also applies the improved BP neural network model to predict the inventory level of an automotive parts company. The results show that the improved algorithm not only significantly exceeds the standard algorithm but also outperforms some other improved BP algorithms both on convergence rate and prediction accuracy.

Prediction Models for Railway Freight Volume Based on Artificial Neural Networks

2014 ◽  
Vol 543-547 ◽  
pp. 2093-2098 ◽  
Author(s):  
Yan Sun ◽  
Mao Xiang Lang ◽  
Dan Zhu Wang ◽  
Lin Yun Liu
Keyword(s):  
Neural Network ◽  
Network Model ◽  
Neural Network Model ◽  
Prediction Accuracy ◽  
Freight Transport ◽  
Generalized Regression Neural Network ◽  
Railway Freight ◽  
Freight Volume ◽  
Artificial Neural ◽  
Generalized Regression

The current China railway freight transport has always been faced with the situation of limited transport resources. Many relative studies have been done to solve the problem of resource shortage. And railway freight volume prediction is the basis of all these studies. With accurate volume prediction, railway freight transport administrations can precisely allocate the transport resources, such as wagons and locomotives. In order to overcome the limitations of traditional prediction methods, in this study, we design four artificial neural network models for prediction, including BP neural network model, linear neural network model, RBF neural network model and generalized regression neural network model. The results of simulation and comparison show that all these models can reach high prediction accuracy and generalized regression neural network has both higher prediction accuracy and better curve fitting capacity compared with other models.

scholarly journals Re-emerging Infectious Disease (RED) Alert tool

2018 ◽  
Vol 10 (1) ◽  
Author(s):  
Maneesha Chitanvis ◽  
Ashlynn Daughton ◽  
Forest M Altherr ◽  
Geoffery Fairchild ◽  
William Rosenberger ◽  
...  
Keyword(s):  
Public Health ◽  
Infectious Disease ◽  
Infectious Diseases ◽  
Disease Incidence ◽  
Emerging Infectious Disease ◽  
Control Measures ◽  
Public Health Response ◽  
Incidence Trends ◽  
Effective Public Health ◽  
Emergence Event

Objective: Although relying on verbal definitions of "re-emergence", descriptions that classify a “re-emergence” event as any significant recurrence of a disease that had previously been under public health control, and subjective interpretations of these events is currently the conventional practice, this has the potential to hinder effective public health responses. Defining re-emergence in this manner offers limited ability for ad hoc analysis of prevention and control measures and facilitates non-reproducible assessments of public health events of potentially high consequence. Re-emerging infectious disease alert (RED Alert) is a decision-support tool designed to address this issue by enhancing situational awareness by providing spatiotemporal context through disease incidence pattern analysis following an event that may represent a local (country-level) re-emergence. The tool’s analytics also provide users with the associated causes (socioeconomic indicators) related to the event, and guide hypothesis-generation regarding the global scenario.Introduction: Definitions of “re-emerging infectious diseases” typically encompass any disease occurrence that was a historic public health threat, declined dramatically, and has since presented itself again as a significant health problem. Examples include antimicrobial resistance leading to resurgence of tuberculosis, or measles re-appearing in previously protected communities. While the language of this verbal definition of “re-emergence” is sensitive enough to capture most epidemiologically relevant resurgences, its qualitative nature obfuscates the ability to quantitatively classify disease re-emergence events as such.Methods: Our tool automatically computes historic disease incidence and performs trend analyses to help elucidate events which a user may considered a true re-emergence in a subset of pertinent infectious diseases (measles, cholera, yellow fever, and dengue). The tool outputs data visualizations that illustrate incidence trends in diverse and informative ways. Additionally, we categorize location and incidence-specific indicators for re-emergence to provide users with associated indicators as well as justifications and documentation to guide users’ next steps. Additionally, the tool also houses interactive maps to facilitate global hypothesis-generation.Results: These outputs provide historic trend pattern analyses as well as contextualization of the user’s situation with similar locations. The tool also broadens users' understanding of the given situation by providing related indicators of the likely re-emergence, as well as the ability to investigate re-emergence factors of global relevance through spatial analysis and data visualization.Conclusions: The inability to categorically name a re-emergence event as such is due to lack of standardization and/or availability of reproducible, data-based evidence, and hinders timely and effective public health response and planning. While the tool will not explicitly call out a user scenario as categorically re-emergent or not, by providing users with context in both time and space, RED Alert aims to empower users with data and analytics in order to substantially enhance their contextual awareness; thus, better enabling them to formulate plans of action regarding re-emerging infectious disease threats at both the country and global level.

scholarly journals Urgensi Pembentukan Aturan Terkait Pencegahan Covid-19 di Indonesia

2020 ◽  
Vol 12 (01) ◽  
pp. 59-70
Author(s):  
Dalinama Telaumbanua
Keyword(s):  
Public Health ◽  
Infectious Disease ◽  
Infectious Diseases ◽  
Government Regulation ◽  
Preventive Measures ◽  
Public Health Emergency ◽  
Legal Certainty ◽  
Government Regulations ◽  
Contagious Disease ◽  
Health Regulations

Covid-19 is a contagious disease that has the potential to cause a public health emergency. Therefore, preventive measures against these types of infectious diseases are mandatory as soon as possible. Indonesia as a nation of law, the prevention of infectious diseases is mandatory to be formed in a rule or regulation. The urgency of forming rules related to the prevention of Covid-19 is obliged to be formed in government regulation and regulation of the Minister of Health because both regulations are the implementation rules of Law No. 6 of 2018 concerning Health. Based on the author's analysis, there are 5 government regulations that must be established in order to perform countermeasures and prevention of infectious disease threats such as Covid-19 and there are 11 mandatory ministerial health regulations that are required to be established In anticipation of the Covid19 threat. Both types of regulations are very useful in anticipating health emergency that ultimately leads to the health of Indonesian people. It is expected that both of rules can be made immediately in order to give legal certainty in preventing the spread of Covid-19 widely. Keyword: Forming Rules, Management, Covid-19   Abstrak Covid-19 merupakan penyakit menular yang berpotensi menimbulkan kedaruratan kesehatan masyarakat. Oleh sebab itu, tindakan pencegahan terhadap jenis penyakit menular tersebut wajib dilakukan secepat mungkin. Indonesia sebagai negara hukum, maka pencegahan terhadap jenis penyakit menular tersebut wajib dibentuk dalam sebuah aturan atau regulasi. Urgensi pembentukan aturan terkait dengan pencegahan Covid-19 ini wajib dibentuk dalam Peraturan Pemerintah dan Peraturan Menteri Kesehatan karena kedua peraturan tersebut merupakan peraturan pelaksanaan daripada Undang-Undang Nomor 6 Tahun 2018 tentang Kekarantinaan Kesehatan. Berdasarkan analisis penulis, ada 5 Peraturan Pemerintah yang wajib dibentuk dalam rangka melakukan tindakan penanggulangan dan pencegahan ancaman penyakit yang mudah menular seperti Covid-19 dan ada 11 Peraturan Menteri Kesehatan terkait yang wajib dibentuk dalam rangka mengantisipasi ancaman Covid-19. Kedua jenis peraturan tersebut sangat berguna dalam hal mengantisipasi kedaruratan kesehatan yang pada akhirnya menjurus pada kekarantinaan kesehatan masyarakat Indonesia. Kiranya kedua jenis peraturan ini segera dibuat dalam rangka memberi kepastian hukum dalam mencegah menularnya Covid-19 secara meluas. Kata Kunci: Pembentukan Aturan, Penanggulangan, Covid-19

Sign in / Sign up

Export Citation Format

Share Document