Phenomenological forecasting of disease incidence using heteroskedastic Gaussian processes: A dengue case study

AbstractThis study describes the development of an integrated dengue alert system (InfoDengue), operating initially in the city of Rio de Janeiro, Brazil. It is a project developed as a partnership between academia and the municipal health secretariat. At the beginning of each epidemiological week, the system captures climate time series, dengue case reporting and activity on a social network. After data pre-processing, including a probabilistic correction of case notification delay, and calculation of dengue's effective reproductive number, indicators of dengue transmission are coded into four dengue situation levels, for each of the city's ten health districts. A risk map is generated to inform the public about the week's level of attention and the evolution of the disease incidence and suggest actions. A report is also sent automatically to the municipality's situation room, containing a detailed presentation of the data and alert levels by health district. The preliminary analysis of InfoDengue in Rio de Janeiro, using historical series from 2011 to 2014 and prospective data from January to December 2015, indicates good degree of confidence and accuracy. The successful experience in the city of Rio de Janeiro is a motivating argument for the expansion of InfoDengue to other cities. After a year in production, InfoDengue has become a unique source of carefully curated data for epidemiological studies, combining epidemological and environmental variables in unprecedented spatial and temporal resolutions.Ethical committee approval: 26910214.7.0000.5240

Download Full-text

Non-Linear Choice Modelling with Gaussian Processes: A Case Study of Neighbourhood Choice in Stockholm

SSRN Electronic Journal ◽

10.2139/ssrn.2869215 ◽

2016 ◽

Author(s):

Richard P. Mann ◽

Viktoria Spaiser ◽

Lina Hedman ◽

David J.T. Sumpter

Keyword(s):

Gaussian Processes ◽

Choice Modelling ◽

Non Linear

Download Full-text

MANAGEMENT TOOLS FOR NETWORK AND PATENT ANALYSIS FOR HEALTH CARE – DENGUE CASE STUDY

Proceedings of 10th CONTECSI International Conference on Information Systems and Technology Management ◽

10.5748/9788599693094-10contecsi/ps-126 ◽

2013 ◽

Author(s):

Jorge L Magalhães ◽

Luc Quoniam

Keyword(s):

Health Care ◽

Patent Analysis ◽

Dengue Case ◽

Management Tools

Download Full-text

Hierarchical Gaussian Processes and Mixtures of Experts to Model COVID-19 Patient Trajectories

10.1101/2021.10.01.462821 ◽

2021 ◽

Author(s):

Sunny Cui ◽

Elizabeth Yoo ◽

Didong Li ◽

Krzysztof Laudanski ◽

Barbara Engelhardt

Keyword(s):

Gaussian Processes ◽

Large Scale ◽

Predictive Performance ◽

Study Patient ◽

Nonparametric Model ◽

Clinical Markers ◽

Spatiotemporal Models ◽

Time Systems ◽

Gp Model

Gaussian processes (GPs) are a versatile nonparametric model for nonlinear regression and have been widely used to study spatiotemporal phenomena. However, standard GPs offer limited interpretability and generalizability for datasets with naturally occurring hierarchies. With large-scale, rapidly-updating electronic health record (EHR) data, we want to study patient trajectories across diverse patient cohorts while preserving patient subgroup structure. In this work, we partition our cohort of over 2000 COVID-19 patients by sex and ethnicity. We develop and apply a hierarchical Gaussian process and a mixture of experts (MOE) hierarchical GP model to fit patient trajectories on clinical markers of disease progression. A case study for albumin, an effective predictor of COVID-19 patient outcomes, highlights the predictive performance of these models. These hierarchical spatiotemporal models of EHR data bring us a step closer toward our goal of building flexible approaches to capture patient data that can be used in real-time systems.

Download Full-text

When time-dependence in disease outcome risk is not captured by impact evaluation modeling studies: a measles vaccination case study

10.1101/2021.07.12.21260376 ◽

2021 ◽

Author(s):

Allison Portnoy ◽

Yuli Lily Hsieh ◽

Kaja Abbas ◽

Petra Klepac ◽

Heather Santos ◽

...

Keyword(s):

Vaccine Coverage ◽

Disease Incidence ◽

Estimation Methods ◽

Model Parameters ◽

Time Varying ◽

Modeling Studies ◽

Measles Vaccination ◽

Secondary Outcomes ◽

The Impact

Background: In modeling studies that evaluate the effects of health programs, the risk of secondary outcomes attributable to infection can vary with underlying disease incidence. Consequently, the impact of interventions on secondary outcomes would not be proportional to incidence reduction. Here we use a case study on measles vaccine program to demonstrate how failure to capture this non-linear relationship can lead to over- or under-estimation. Methods: We used a published model of measles CFR that depends on incidence and vaccine coverage to illustrate the effects of: (1) assuming higher CFR in 'no-vaccination' scenarios; (2) time-varying CFRs over the past; and (3) time-varying CFRs in future projections on measles impact estimation. We evaluated how different assumptions on vaccine coverage, measles incidence, and CFR levels in 'no-vaccination' scenarios affect estimation of future deaths averted by measles vaccination. Results: Compared to constant CFRs, aligning both 'vaccination' and 'no-vaccination' scenarios with time variant measles CFR estimates led to larger differences in mortality in historical years and lower in future years. Conclusions: To assess consequences of interventions, impact estimates should consider the effect of 'no-intervention' scenario assumptions on model parameters to project estimated impact for alternative scenarios according to intervention strategies and investment decisions.

Download Full-text

Assessing losses related to actual disease incidence in a region's population: an economic aspect (a case study of the Altai region of russia)

Health Risk Analysis ◽

10.21668/health.risk/2013.2.09.eng ◽

2013 ◽

pp. 73-78

Author(s):

A.A. Ushakov ◽

◽

I.P. Saldan ◽

O.I. Goleva ◽

T.N. Karpova ◽

...

Keyword(s):

Disease Incidence ◽

Economic Aspect ◽

Altai Region

Download Full-text

Assessing the effects of varied temperature and pH on the growth and sclerotial formation of Rhizoctonia solani Kuhn, isolated from paddy field: a case study.

International Journal of Life Sciences ◽

10.3126/ijls.v8i2.10226 ◽

2014 ◽

Vol 8 (2) ◽

pp. 4-9

Author(s):

Sambit Dutta ◽

Souvik Das ◽

Abhijit Sarkar ◽

Jayanta Tarafdar ◽

Ashim Chowdhury

Keyword(s):

Rhizoctonia Solani ◽

Abiotic Factors ◽

Disease Incidence ◽

Tropical Asia ◽

Factors Associated ◽

Sheath Blight Disease ◽

Varied Temperature ◽

Sclerotial Formation ◽

Blight Disease

Rhizoctonia solani Kuhn, is one of the most important pathogen of rice causing the sheath blight disease which accounts for heavy crop losses in all the major rice growing areas of the world, including tropical Asia. The proliferation of this soil borne pathogen depends on many abiotic factors. The two most common factors associated with the growth and spread of the pathogen is pH and temperature which has been studied in the present study. Both pH and temperature had significant effect on the growth of the fungal mycelial and sclerotial formation. A non-linear regression model for growth of the fungal mycelia revealed that growth occurred best at pH 6.0 and temperature 30°C for all treatments studied. No growth was observed below 15°C. In agricultural fields sclerotia, forms the means for the spread of the pathogen causing secondary disease incidence as well as survival of the pathogen in harsher conditions. Sclerotia formed between the temperatures of 20°C and 30°C, with an optimum at 30°C and pH 6.0. DOI: http://dx.doi.org/10.3126/ijls.v8i2.10226 International Journal of Life Sciences Vol.8(2): 2014; 4-9

Download Full-text

A cassava mosaic disease epidemiologic factors analysis in the Bas-Congo (DR Congo) zone of savannah. case study: “Secteur de Boko”

International Journal of Biological Research ◽

10.14419/ijbr.v3i1.4069 ◽

2015 ◽

Vol 3 (1) ◽

pp. 56

Author(s):

Bansimba Mukiese J.-Roger

Keyword(s):

Disease Incidence ◽

Mosaic Disease ◽

Cassava Mosaic Disease ◽

Dr Congo ◽

Local Varieties ◽

Epidemiologic Factors ◽

Factors Analysis ◽

Foliar Symptoms ◽

The Mean

<p>This paper aims at evaluating epidemiologic parameters of the African Cassava Viral Mosaic Disease (CMVD) in 25 cassava farmer's fields of «Secteur de Boko» (DR Congo) on the basis of foliar symptoms observation. The results showed that the infection of the local varieties comes mainly from the cuttings whereas that of the improved varieties comes from the flies. The disease incidence varied between 10% and 88.33% with an average of 43.8%. Severity varied between 2.42 and 3.42 with an average of 2.913. The gravity varied between 17.86% and 87.81% while the systemicity varied between 21.42 and 96.66%. The mean number of whiteflies (<em>Bemisia tabaci</em>) by plant (2.068) revealed a preference of the vectors for the improved variety Kindombe which, however, presents a low severity and a less marked gravity. Globally, the CMVD was more severe for the local variety Mpeko with a score of 3.00. A high correlation was found between the gravity and the number of neighbour fields (-0,437*), the density of culture (-0,431*) and also the systemicity of the disease (0,779**). In addition; it appeared that the land topography strongly influences the disease severity (-0,542**).</p>

Download Full-text