Using data-driven approaches to model tick bites in Dutch gardens

Tick presence in the Netherlands (Swart et al., 2014)

Objective:

This MSc topic will use machine learning to model the occurrence of tick bites in Dutch gardens. This model will be based on volunteered observations coupled with environmental and cadastral data that allow characterizing different types of gardens.
Tick questing

Description:

Volunteered geographic information (VGI) is becoming an important alternative and/or complementary data source where there is a lack on data at appropriate spatial and/or temporal resolutions. For instance, to study vector-borne diseases.

In vector-borne diseases, a host (usually a parasite, like a tick) jumps between individuals (often from different species) to spread the disease. In The Netherlands, there is a network of volunteers that collect data about tick bites. Some of these ticks are sent to the national public health institute (RIVM) to be checked for Lyme disease. It is well known that the abundance of ticks depends on environmental factors, such as type of vegetation, temperature or humidity, and the presence of host animals. Some of this environmental data that can be obtained by analyzing remote sensing images, weather data and monitoring the presence of wild animals (deer or rodents) that act as intermediate hosts for the ticks.

This MSc topic will examine the tick bites that occurred in gardens (tick radar (website in Dutch). These volunteered observations will be integrated with environmental and cadastral information to determine whether certain garden topologies are preferred by ticks. For this, the MSc candidate will first perform a detailed exploratory data analysis and then s/he will create a model that, fed with the volunteered observations and environmental and ancillary datasets, will model the occurrence of tick bites in Dutch gardens. This research contributes to determine the suitability of volunteer data to study complex public health issues.

One PhD student is currently working on this topic (Irene Garcia). She will advise and co-supervise the MSc candidate. An analytical mind and affinity with scripting are needed in order to successfully complete this research.

References:

  • Mulder, S., van Vliet, A.J.H, Bron, W.A. Gassner, F. & Takken, W. (2013). High risk of tick bites in Dutch gardens. Vector-Borne and Zoonotic Diseases, 13(12): p. 865-871.

  • Zeman, P. & C. Benes, Peri-urbanisation, counter-urbanisation, and an extension of residential exposure to ticks: A clue to the trends in Lyme borreliosis incidence in the Czech Republic? Ticks and tick-borne diseases, 2014. 5(6): p. 907-916.

  • Swart, A., Ibañez-Justicia, A., Buijs, J., van Wieren, S.E, Hofmeester, T.R., Sprong, H. & Takumi, K. (2014). Predicting tick presence by environmental risk mapping. Frontiers in Public Health: Vol. 2, Article 238.

Domain(s):

Study Program(s):

Researchers working on this field: