Richard Kiang

1
Malaria Modeling for Thailand Korea
NASA Techniques and Call for Validation Partners

• Richard Kiang
• NASA Goddard Space Flight Center
• Greenbelt, MD 20771

2
Acknowledgement
AFRIMS Dr. Jame Jones WRAIR Dr. Russell
Coleman Dr. R. Sithiprasasna Dr. Gabriella
Zollner USU Dr. Donald Roberts NDVECC Dr.
David Claborn Dr. Richard Andre Dr. Leon
Robert Ms. Penny Masuoka NGA Mr.
John Doty DOS Mr. Andrew Herrup UC
Davis Dr. John Edman Cornell Univ. Dr. Laura
Harrington Mahidol Univ. Dr. S.
Looareesuwan Thai MOPH Dr. J. Sirichaisinthop D
r. P. Singhasivanon Mr. S. Nutsathapana Dr.
S. Leemingsawat Dr. C. Apiwathnasorn
RTSD Gen. Ronnachai Thai Army Lt. P.
Samipagdi Dr. Kanok
3
Mekong Malaria Filariasis
Kanchanaburi
Malaria Cases
Ikonos
Source SEATMJ
Ban Kong Mong Tha
Filariasis poster
Field work / Mahidol
Field work / AFRIMS
Richard.Kiang_at_nasa.gov
4
Mekong Malaria and Filariasis

• MODELS
• Vector Habitat Model
• Malaria Transmission Model
• Risk Prediction Model

• temperature
• precipitation
• humidity
• surface water
• - wind speed direction
• land cover
• vegetation type
• transportation network
• - population density

Data

• MEASUREMENTS
• Ikonos
• ASTER
• Landsat
• MODIS
• etc.

Richard.Kiang_at_nasa.gov
5
INTEGRATED PEST MANAGEMENT FOR DOD
PROJECT OBJECTIVES
6
Objectives, Approaches Preliminary Results
Identifying key factors that sustain or intensify
transmission
Habitat identification
Risk prediction
Textural-contextual classifications significantly
increase landcover mapping accuracy using
high resolution data such as Ikonos.
Nonparametric model computes the likelihood of
disease outbreak using meteorological and
epidemiological time series as input.
Discrete Wavelet Transform is used to
differentiate confusion vegetation types.
Wavelet Transform and Hilbert-Huang Transform
Empirical Mode Decomposition identify the driving
variables that lead to disease outbreaks and
provide more accurate predictions.
Evaluated Thail military airborne data and
established neural network rectification
capability.
Richard.Kiang_at_nasa.gov
7
(No Transcript)
8
Bamboo Cups Kanchanaburi
9
Washington, D.C.
Space Imagings Ikonos imagery
10
Steps in Performing Discrete Wavelet Transform
approx
down sample cols
down sample rows
low pass on rows
low pass on cols
vertical edges
image
high pass on rows
high pass on cols
horizontal edges
diagonal edges
11
Textural Feature Extraction using Discrete
Wavelet Transform








Approx
H
Horizontal Edges
A square neighborhood in the imagery data
V
D
n-D entropy vector
Diagonal Edges
Vertical Edges
12
Class Separability with Textural Features
extracted by Discrete Wavelet Transform
13
Entropy Derived from DWT as Textural Measure to
Aid Classification
Ikonos
Last 8x8 neighborhood
Its WC from DWT
Combined with panchromatic
Largest entropy
2nd largest entropy
1m resolution
14
North Korea Malaria Transmission
15
Camp Greaves and Surrounding AreaKyunggi, South
Korea
kr4_truecolor_brightened.jpg
Space Imagings Ikonos imagery
16
Pseudo Ground Truth
Kr34_pseudogt.jpg
17
Panchromatic Intensity
(RGB)/3
(NRB)/3
Space Imagings Ikonos imagery
18
  From Cook et al. Ikonos Technical Performance
Assessment 2001 SPIE Proceedings, Algorithms for
Multispectral, Hyperspectral, ..., p.94.  
19
Classification Accuracy using Pan-Sharpened Ikonos
Data ( 1 meter resolution)
20
Detection of Ditches using 1-meter Data(Larval
Habitats of An. sinensis)
21
NDVI from AVHRR Measurements
NDVI Normalized Difference Vegetation
Index AVHRR Advanced Very High Resolution
Radiometer

• ? NDVI (near infrared red)
• (near infrared red)
• ? Can be used to infer ground cover
• and rainfall.
• Can be derived from other sensors
• as well.

Compiled by NOAA/NESDIS for Feb. 13, 2001
22
Post-Processing with Class Frequency Filters
 
23
Sample Image of Royal Thai Survey Departments
Airborne Instrument
? From a Beechcraft B200 Super King Air ?
Effective surface resolution approx. 1.5m
24
Using Neural Networkto Rectify
AircraftMeasurements
Simulated Measurements Generated by Scanner Model
Rectified
open squares real positions shaded squares
fitted positions
25
Ban Kong Mong Tha Sanghlaburi, Kanchanaburi,
Thailand
26
Anopheles dirus forest shaded pools hoofprints in or at the edge of forests with increasing deforestation, adapting to orchards, tea, rubber and other plantations.
An. minimus forest fringe flowing waters (foothill streams, springs, irrigation ditches, seepages, borrow pits, rice fields) shaded areas grassy and shaded banks of stable, clear, slow moving streams.
An. maculatus seepage waters streams pools pond edges ditches and swamps with minimal vegetation sunlit areas.
27
An. dirus
An. minimus
28
(No Transcript)
29
TRANSMISSION MODEL
Microepidemiology Data
Satellite Meteor. Data
Local Environment
Landcover
Vector Ecology
Population Database
Vector Control
Host Behaviors
Dwelling
Medical Care
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
Landsat TM Image over Mae La
35
Mae La Camp
36
(No Transcript)
37
Sources CDC DVBID Rutgers Univ.
Entomology Dept./NJMCA
38
Airborne Remote Sensing
In late 19th Century
ER-2 Fleet
39
Neural Network Classification of GER 63-channel
Scanner Data
Architecture Training Acc. Rel. Classif. Acc.
1 hidden layer with 1 node 88.41 85.52
1 hidden layer with 3 nodes 99.07 97.93
1 hidden layer with 5 nodes 98.86 97.52
2 hidden layers each with 3 nodes 99.07 97.62
2 hidden layers each with 5 nodes 99.38 97.83
 
40
1985-1999 SIESIP ½½ temp, precip
2000-2003 SIESIP ½½ temp, precip
1985-2003 NCEP 2½2½ rel. humidity
1998-2003 TRMM ½½ precip
1999-2003 MODIS 55 km² surface temp, lifted
index, moist., etc.
1985-2000 AVHRR PF 88 km² NDVI
1999-2003 MODIS 88 km² NDVI
41
Time-Frequency Decompositions
Dengue Cases Kuala Lumpur
Fourier Transform
Hilbert-Huang Transform
Wavelet Transform
42
RISK PREDICTION MODEL
Nonparametric model computes the likelihood of
disease outbreak using meteorological and
epidemiological time series as input.
Wavelet Transform and Hilbert-Huang Transform
Empirical Mode Decomposition identify the driving
variables that lead to disease outbreaks and
provide more accurate predictions.
43
NASA Goddard Space Flight Center
Space Imagings Ikonos imagery
Landsat-1 MSS
44
NASA/GSFC Close-Up
Pan 1m MS 4m
Space Imagings Ikonos imagery
45
2-Year Prediction of Malaria Cases Based on
Environmental Parameters (temperature,
precipitation, humidity, vegetation
index) Ratchaburi, Thailand
46
Ban Kong Mong Tha Sanghlaburi, Kanchanaburi,
Thailand