Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2012 Nov 5:12:124.
doi: 10.1186/1472-6947-12-124.

A data-driven epidemiological prediction method for dengue outbreaks using local and remote sensing data

Anna L Buczak  1 , Phillip T Koshute, Steven M Babin, Brian H Feighner, Sheryl H Lewis
Affiliations

A data-driven epidemiological prediction method for dengue outbreaks using local and remote sensing data

Anna L Buczak et al. BMC Med Inform Decis Mak. .

Abstract

Background: Dengue is the most common arboviral disease of humans, with more than one third of the world's population at risk. Accurate prediction of dengue outbreaks may lead to public health interventions that mitigate the effect of the disease. Predicting infectious disease outbreaks is a challenging task; truly predictive methods are still in their infancy.

Methods: We describe a novel prediction method utilizing Fuzzy Association Rule Mining to extract relationships between clinical, meteorological, climatic, and socio-political data from Peru. These relationships are in the form of rules. The best set of rules is automatically chosen and forms a classifier. That classifier is then used to predict future dengue incidence as either HIGH (outbreak) or LOW (no outbreak), where these values are defined as being above and below the mean previous dengue incidence plus two standard deviations, respectively.

Results: Our automated method built three different fuzzy association rule models. Using the first two weekly models, we predicted dengue incidence three and four weeks in advance, respectively. The third prediction encompassed a four-week period, specifically four to seven weeks from time of prediction. Using previously unused test data for the period 4-7 weeks from time of prediction yielded a positive predictive value of 0.686, a negative predictive value of 0.976, a sensitivity of 0.615, and a specificity of 0.982.

Conclusions: We have developed a novel approach for dengue outbreak prediction. The method is general, could be extended for use in any geographical region, and has the potential to be extended to other environmentally influenced infections. The variables used in our method are widely available for most, if not all countries, enhancing the generalizability of our method.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Dengue cases per week.
Figure 2
Figure 2
(A) Departments in Peru. We used Loreto (shaded in yellow) for our study. (B) Fifty-one districts in the Loreto department. We used six districts for analysis (shaded in darker colors).
Figure 3
Figure 3
Weekly dengue Incidence Rate per 1000 residents.
Figure 4
Figure 4
4-Week dengue Incidence Rate per 1000 residents.
Figure 5
Figure 5
Example of satellite-derived daily rainfall for Peru. Units are mm.
Figure 6
Figure 6
Example of dividing a grid cell into subcells and counting the centroids falling within each district.
Figure 7
Figure 7
Example of day-time temperature data for a given 8-day interval. Units are degrees Celsius. Procedures were developed in MATLAB to remove missing data. The dark blue spots near Iquitos (corresponding to a temperature near 0 C) were removed since they were corresponding to missing data.
Figure 8
Figure 8
Illustration of spatial resolution of different variables.
Figure 9
Figure 9
Example of NDVI values for a given 16-day interval.
Figure 10
Figure 10
Example of EVI values for a given 16-day interval.
Figure 11
Figure 11
Dengue prediction method developed.
Figure 12
Figure 12
Membership functions for the fuzzy variable Temperature: Low, Normal, Fever, Strong Fever and Hyperthermia.
Figure 13
Figure 13
Membership functions for the variable NDVI _T-8_T-4 as defined in the JHU/APL FARM software. The membership functions are: Med, High and Very High.
Figure 14
Figure 14
Membership functions for the variable Sanitation as defined in the JHU/APL FARM software. The membership functions are: Low, Med and High.
Figure 15
Figure 15
Membership functions for the variable SSTA_T-2 as defined in the JHU/APL FARM software. The membership functions are: Low, Med and High.
Figure 16
Figure 16
4-Week prediction results for the test set.
Figure 17
Figure 17
4-Week prediction results for the training set. When the blue curve (actual) falls within the yellow bar (prediction) there is no error. When the blue curve falls outside of the yellow bar, there is a prediction error.
Figure 18
Figure 18
4-Week prediction results for the validation set.
Figure 19
Figure 19
Prediction on the training data for Iquitos. The predicted Low is shown in green, the predicted High in red, and the actual values are shown in blue. The gaps with no values correspond to data that was not used in training. When the blue curve (actual) falls within the green bar (predicted Low) or red bar (predicted High) there is no error. When the blue curve falls outside of those bars, there is a prediction error.
Figure 20
Figure 20
Prediction on the validation data for Iquitos. The predicted Low is shown in green, the predicted High in red, and the actual values are shown in blue. The gaps with no values correspond to data that was not used in validation.
Figure 21
Figure 21
Prediction on Test data for Iquitos. The predicted Low is shown in green, the predicted High in red, and the actual values are shown in blue. The gaps with no values correspond to data that was not used in testing.
Figure 22
Figure 22
Comparison of LR and FARM-based results for weekly predictions three-weeks in advance (T+3).
Figure 23
Figure 23
Comparison of LR and FARM-based results for weekly predictions four-weeks in advance (T+4).
Figure 24
Figure 24
Comparison of LR and FARM-based results for four week interval prediction: weeks T+4 through T+7.
See this image and copyright information in PMC

References

    1. Focks DA, Daniels E, Haile DG, Keesling JE. A simulation model of the epidemiology of urban dengue fever: literature analysis, model development, preliminary validation, and samples of simulation results. Am J Trop Med Hyg. 1995;53(5):489–506. - PubMed
    1. Gibbons RV, Vaughn DW. Dengue: an escalating problem. BMJ. 2002;324(7353):1563–1566. doi: 10.1136/bmj.324.7353.1563. - DOI - PMC - PubMed
    1. TDR/WHO. Dengue Guidelines for Diagnosis, Treatment, Prevention and Control. Geneva: World Health Organization; 2009. Available: http://whqlibdoc.who.int/publications/2009/9789241547871_eng.pdf. Accessed 26 October 2012. - PubMed
    1. Ranjit S, Kissoon N. Dengue hemorrhagic fever and shock syndromes. Pediatr Crit Care Med. 2010;12(1):90–100. - PubMed
    1. Guzman MG, Kouri G. Dengue: an update. Lancet Infect Dis. 2002;2:33–42. doi: 10.1016/S1473-3099(01)00171-2. - DOI - PubMed

Publication types