Navigation

Digital Health Geographies Research Group (Walker)

How do digital geographical data foster innovative understandings of space, place, and health?

How can we reconcile spatial analysis and machine learning with the need for local knowledge and situatedness in health/medical geographical research?

These are the key questions guiding our thinking and research, as pursued in the range of projects listed below.

The rapid nascence of data availability, powerful algorithms, and quantitative methodology are necessitating novel conceptualisations of “digital”, serving both as an applied means of constructing information and as an arena for critical inquiry into emergent modes of representation. With a broad focus on the interplays between human health and social/built/natural environments, the Digital Health Geographies Research Group explores the confluence of applied machine learning and the analysis of situated, community-scale sociocultural practice through empirical study.

Can artificial intelligence algorithms identify community-scale socioeconomic and built-environment risk factors?

The project examines geographical associations between COVID-19, socioeconomic status, and the built/natural environment. Using contemporary spatial epidemiological techniques and Bayesian machine learning we are presently testing novel methodological combinations to triangulate factors underlying the spatial and temporal structures of the epidemic in Germany.
In recent years artificial intelligence and machine learning algorithms have raised attention and critique. These so-called “black-box” models yield impressive predictive performance when analysing high-dimensional, complex datasets, but at the cost of interpretability. Using COVID-19 in Germany as an example, the Digital Health Geographies Research Group seeks to explore ways to open up these “black boxes” and dissect their algorithms, with the ultimate goal of developing methods for implementing prior geographical knowledge in machine learning.

External Partners:

  • Christopher Scarpone, Ryerson University, Toronto, Canada.

Theses:

  • Anna Stadlmeier, MA Thesis: Social connectedness in the context of COVID-19

Publications:

  • Scarpone C., Brinkmann S.T., Große T., Sonnenwald D., Fuchs M., Walker BB. A multimethod approach for county-scale geospatial analysis of emerging infectious diseases: a cross-sectional case study of COVID-19 incidence in Germany. Int J Health Geogr 19, 32 (2020). https://doi.org/10.1186/s12942-020-00225-1.

Violência no Brasil: análise de comunidades e espaço / Violence in Brazil: Analysis of Community Environments

Brazil is known worldwide for its vibrant cultural diversity and beautiful landscapes. However, increasing awareness of deeply embedded  social issues such as social inequality, poverty, and violence underscores the need for more nuanced understandings of their interrelations with space and place. VIBRANCE is a collaborative study bringing together researchers from Brazil, Germany, and Canada to develop and implement innovative interdisciplinary approaches to studying space, place, and violence. By using advanced quantitative techniques and situated qualitative inquiry, VIBRANCE assesses various domains of social and spatial risk factors of violence at multiple scales across the country, seeking to advance understandings of the complex geographies of violence and to identify policies and programmes to reduce the burden of violence in Brazilian communities.

External Partners:

  • Dr. Kevan Guilherme Nóbrega Barbosa, Centro Universitário CESMAC, Macieó, Brazil.

Theses:

  • Anton Leutner, BA Thesis: Spatialities of Violence: Analysis socioeconomic patterns of gun violence in Rio de Janeiro and Recife, Brazil
  • Antonia Bauer, BA Thesis: Instrumentalization of space through violence
  • Julian Hofmann, BA Thesis: Gemeinwohlökonomie im brasilianischen Kontexten

How do (un)informed discourses about pandemics and space develop and (d)evolve through digital (social)media?

Using COVID-19 as an example, the Digital Health Geographies research group uses automated knowledge mining methods to investigate, on the basis of different types of text, (1) which spatial-cognitive strategies are used to consistently locate crises outside ones living environment, (2) which events modify these narratives over time and (3) how newly emerging arguments discursively re-bind these events. The aim of the study is to alleviate the socioeconomic consequences of crises by revealing the suspensive character of certain narratives for necessary political decision-making processes.

Prospective Urban and Rural Epidemiology: Spatial Analysis of Cardiovascular Environments

This project seeks to disentangle the complex relationships between socioeconomic status, the built/natural environment, and chronic diseases, specifically focussing on cardiovascular disease and diabetes mellitus risk factors. Using epidemiological techniques and machine learning, we develop socioeconomic and environmental indices to estimate the spatial component of CVD risk. This project is being conducted as part of the Prospective Urban and Rural Epidemiology (PURE) Study.

External Partners:

  • Prof. Scott A. Lear, Providence Health, Vancouver, Canada

Spatial Transcoding of Artificial Geographical Environments

Open Worlds, Digital Landscapes: digital inscriptions of cultural mythologies in playable open worlds and built environments.

In the field of Place Research, visual spaces, both real-world and artificial, provide a stage for shared social practice. There exist different forms of visual spaces, ranging from complete virtual worlds over augmented reality to geoservices, which add digital information on site.
Individual practice is bound to the participant’s specific experiences and cultural context. As it can be revealed by the metaphor of game and play, we explore how these aspects are represented, expressed, and enacted in different game environments. Derived questions cover: How is space used as an element within the game to create a flow state, which binds the player to the game?

Theses:

  • Daniel Sonnenwald, BA Thesis: “A galaxy far far away…”:  Space Syntax, design and cinematography in the interactive medium of video games.

On the basis of a selected example, the author investigates the functionality of space in videogames. How is space used as an element within the game to create a flow state, which binds the player to the game? How is affordance created and communicated trough digital architecture? In the eyes of the author, a combination of approaches from grounded theory and the analysis of space syntax is the means of choice to investigate these questions.

  • Lukas Suk, BA Thesis: Roles of Landscape in Red Dead Redemption 2
  • Bryan Maland, BA Thesis: Geographische Analyse des Videospiels Red Dead Redemption 2 mithilfe der Prospect Refuge Theory

  • Marika Cordes, MA Thesis: Global South-North Innovation Mobilities in Medical Diagnostic Technologies – A Case Study of Rats for Tuberculosis Detection.
  • Lasse Harkort, BA Thesis: Analysis of Collisions with Cyclist Participation by Regional Typology in Germany

2020
  • Scarpone C., Brinkmann S.T., Große T., Sonnenwald D., Fuchs M., Walker BB. A multimethod approach for county-scale geospatial analysis of emerging infectious diseases: a cross-sectional case study of COVID-19 incidence in Germany. Int J Health Geogr 19, 32 (2020). https://doi.org/10.1186/s12942-020-00225-1.
  • Walker BB, Shashank A, Gasevic D, Schuurman N, Poirier P, Teo K, Rangarajan S, Yusuf S, Lear SA. The Local Food Environment and Obesity: Evidence from Three Cities. Obesity. 2020; 28(1):40-45. DOI: 10.1002/oby.22614.
  • Schuurman N, Walker BB, Swanlund D, Amram O, Yanchar N. Qualitative field observation of pedestrian injury hotspots: A mixed-methods approach for developing built- and socioeconomic-environmental risk signatures. IJERPH. 2020; 17:2006-2021. DOI: 10.3390/ijerph17062066.
  • Wister A, Shashank A, Rosenkranz L, Walker BB ,Schuurman N. Multimorbidity and socioeconomic deprivation among older adults: a cross-sectional analysis in five Canadian cities using the CLSA. Journal of Aging and the Environment. 2020. DOI: 10.1080/26892618.2020.1734138.
  • Ho HC, Man HY, Wong MS, Shi Y, Walker BB. Perceived differences in the (re)production of environmental deprivation between sub-populations: a study combining citizens‘ perceptions with remote-sensed and administrative data. Building and Environment. 2020. DOI: 10.1016/j.buildenv.2020.106769.
2019
  • Walker BB, Canham S, Wister A, Fang ML. Disparities in local access to culturally-relevant residential care and assisted living facilities: a GIS-facilitated study of East Asian seniors’ care in Vancouver, Canada. Journal of Care for the Elderly. 2019; 1540:353X.
  • Enoguanbhor E, Lakes T, Nielsen JØ, Gollnow F, Walker BB. Land cover change detection in the Abuja City-Region, Nigeria: Integrating GIS and remotely sensed data to support decision-making in land use planning. Sustainability. 2019; 11:1313. DOI: 10.3390/su11051313.
  • Barbosa K, Walker BB, Ferrira R. Spatial Epidemiological characteristics of interpersonal violence in Campina Grande, Brazil: a comparative geographic study of violence in domestic versus community settings. PLoS ONE. 2019; 14(1):e0208304. DOI: 10.1371/journal.pone.0208304.