Data science has really made its mark in both the research community and the rest of society. This broadly useful research field can even help to make the world a better place, for example by curbing the transmission of malaria in Dar es Salaam and other major cities in Africa.
A particularly invasive species of malaria mosquito native to Asia (Anopheles stephensi) has migrated to many cities in Africa, including Tanzania’s largest city, Dar es Salaam. It is especially fond of the humid environments in the cities. If this species of mosquito really spreads, there is a risk of major outbreaks of malaria – especially in the densely populated informal settlements.
To avoid these outbreaks, an international research team of epidemiologists, biologists, geographers, computer scientists and architects will use data science to study how mosquito reproduction is linked to architecture and urban planning in Dar es Salaam. The research team will develop a special deep learning algorithm that will help local authorities and others in their efforts to combat malaria and other mosquito-borne diseases.
“We have assembled several disciplines so that together we can try to determine how to use data science and images to understand more about malaria mosquitoes and their impact in urban areas,” explains Jakob Brandtberg Knudsen, Professor and Dean – Architecture, Royal Danish Academy, Copenhagen, who is leading the project in Dar es Salaam.
Collecting data – and mosquitoes
The researchers will use high-resolution satellite images and drone images in the project to map the houses in especially the informal settlements of the city – sometimes referred to as slums. The researchers will also collect data on the ground, going from house to house, interviewing residents and collecting mosquito samples.
“Although we use very high-resolution satellite images similar to those on Google Earth, we also have people on the ground who will register the dwellings, interview the people who live there, determine what materials are used to build the houses and find out what the surroundings are like. We will also collect mosquitoes from inside the houses and from the surrounding areas. We are doing this to determine where the mosquitoes are and the associations between the types of dwellings, the surroundings and how many mosquitoes there are – and thus also the risk of developing malaria,” adds Jakob Brandtberg Knudsen.
Jakob Brandtberg Knudsen was previously involved in research that has mapped housing conditions in sub-Saharan Africa, investigating how improving housing conditions affects the risk of getting malaria. The studies showed some fairly direct associations between how well houses are constructed and the prevalence of diseases. The occupants of well-constructed houses had less malaria and other types of diseases, such as respiratory diseases and diarrhoea.
This research was published in 2019 in two articles in Nature: “Mapping changes in housing in sub-Saharan Africa from 2000 to 2015” and “Reduced mosquito survival in metal-roof houses may contribute to a decline in malaria transmission in sub-Saharan Africa”. These results largely comprise the basis for the new research project in Dar es Salaam, especially because the researchers discovered that roof structures strongly influence the spread of malaria mosquitoes. Houses with sheet-metal roofs rather than thatched roofs have fewer mosquitoes because they do not survive as well.
This provided the inspiration for using different types of data and linking them to the new project in Dar es Salaam, which aims to develop a deep learning algorithm that can be trained to recognise different types of buildings and surroundings on the satellite images – such as roofs and ponds – so the images will contain a lot of information that can be quantified and converted into risks of malaria outbreaks.
More data creates optimal basis for decision-making
Once the algorithm has been developed, local authorities, nongovernmental organisations, WHO and others striving to prevent malaria can use it. More information about local conditions will make it easier to implement appropriate preventive measures that can lead to fewer cases of malaria and other diseases.
The project is therefore a good example of what lies at the heart of data science: collecting as much data on a given subject as possible to create the optimal basis for decision-making.
“Data science strives to measure the world more and more accurately – what we call digitalisation. Based on these data, we can build models that can enable evidence-based decisions based on facts rather than on opinion. We are increasingly doing this across all parts of society. This means that data science is evolving extremely rapidly and is increasingly important. In many respects, this field is the locomotive that will drive the economic growth from digitalisation,” says Lars Kai Hansen, Professor and Head of Section, Department of Applied Mathematics and Computer Science, Technical University of Denmark and Chair of the Danish Data Science Academy.
In this episode of the Forskningsfortællinger podcast (in Danish), you can listen to Jakob Brandtberg Knudsen explain more about the research project in Dar es Salaam. Lars Kai Hansen also explains more about the current and future role of data science in society.
