Transcript GeoVisualisation of health data

GeoVisualisation of health
data
Prof Clive Sabel
Geography, CLES and
European Centre for Environment & Human Health
Spatial data mining
case-studies
- Childhood Obesity in Devon
- Neurological disease in Finland
- Road Traffic Accidents in New Zealand
Childhood Obesity in
Devon
US election 2008
‘Distorted’
maps:
Cartograms
US election 2004
HIV Prevalance
Physicians working
Neurological disease in
Finland
Cases MND
• 1000 cases
• 3113 case
residences
• But population
naturally clusters
in cities
• So search for
‘excess’ over and
above the
background rate
Residential Clustering of
Motor Neurone Disease
Finland
• Kernel (Density) Estimates of
All places of residence
• Statistical Significance tested
by Monte-Carlo Simulation
• Note large peak in SE Finland
Space-Time
Visualisation
• Residential Migration
• Animation
• Space: Finland
• Time: 1965 – 1990
• Illustrates Spatial and Temporal
variation
• Note:
• persistent dark areas in SE
• In SW, at varying times, areas
switch from high to low rates
of disease
MND Individual Migration Histories
• Same data as before, but now
showing individual migration
trajectories.
• The larger densities of colour
identify the larger cities
Road Traffic Accidents in
New Zealand
Christchurch, NZ
City Centre
Also a Kernel
density estimate
of RTAs.
Compare model
results (black
polygon) with
observations
(red-blue image)
Note more
accidents than
modelled in the
city centre, and
around junctions
Left:
Temporal analysis of all road accidents
- accidents reducing over time, due to
increased safety measures introduced.
Right:
Temporal analysis of just accidents
involving cyclists
- accidents increasing in the circled area,
why?
Drive Safely!