hul 2004 
7 image 
D. 
f digital 
fication. 
pp. 297- 
analysis 
eie N.F. 
ariables 
Ji of 
)cessing 
Remote 
thermal 
national 
14 -156. 
review 
p. 1461- 
-sensed 
| for the 
levation 
[maging 
elgium, 
ion of 
e Flood 
2nd Int 
ig, ITC 
ital 
& 
ndcover 
4th Int. 
ç, 27-29 
odeling 
tion of 
, Open- 
spatial 
ication. 
dryland 
data. 
Chester 
S 
ling the 
SPATIAL AND TEMPORAL ANALYSIS OF CUTANEOUS LEISHMANIASIS 
INCIDENCE IN SÁO PAULO - BRAZIL 
Cristina Aparicio , Marisa Dantas Bitencourt 
Ecology Department - Biosciences Institute — University of Säo Paulo- Brazil, 
Rua do Matáo, travessa 14, n? 321 — Sáo Paulo / SP - CEP: 05508-900 
crisap@usp.br; tencourt@ib.usp.br 
KEYWORDS: Change Detection, Multitemporal, Landscape, Forestry, Monitoring. 
ABSTRACT 
Cutaneous Leishmaniasis transmission occurs in a complex cycle where the environment has a significant influence. Some important 
parameters related to this cycle can be qualified and quantified with Remote Sensing and Geoprocessing, employed to study 27 cases 
of CL occurred at Itapira municipality (Sáo Paulo/Brazil), reported between 1992 and 1997. Forest remnants related to these cases 
were delimited and the distances between dwellings and remnants were calculated. Contact Risk Zones were defined taking the 
remnants border and expanding it, according to criteria based on the mosquitoes flight range. Also, the influence of vegetation 
density and deforestation on the CL incidence was analyzed through: 1) a Normalized Difference Vegetation Index (NDVI) and 2) a 
Multitemporal Principal Component Analysis (PCA). The relative percentages of NDVI classes found around the remnants showed a 
high Vegetation Index in this region, varying between 0.45 and 1.0 in approximately 50% of remnant surrounding areas. The Second 
Principal Component derived from PCA was used to determine the areas where deforestation seems to have an important role in CL 
incidence. The analysis carried out indicate that three types of transmission may be occurring in that region: (i) an intra-forest 
transmission in locals where a disease focus was localized within a dispersion radius of the mosquito; (ii) an extra-forest 
transmission probably related to the existence of a high percentage of dense vegetation around remnants where the mosquito can 
flight and (iii) transmission in human dwellings, possibly caused by a mosquito domestication process. 
1 - INTRODUCTION 
Remote Sensing and Geoprocessing are becoming very 
important for Landscape Epidemiology, because it enables the 
analysis of some environmental variables related to several 
diseases incidence (Hugh-Jones 1989, Verhasselt 1993, 
Washino and Wood 1994, Clarke et al. 1996, Croner et al. 
1996, Beck et al. 2000). Cutaneous Leishmaniasis (CL) is one 
of these diseases, in which the transmission profile includes 
landscape elements. 
CL is among the six most important infected-parasitic diseases 
of the world, and is the second most important caused by 
protozoan, right below malaria. According to Walsh et al. 
(1993), in South America CL distribution has been influenced 
by zoofilic vectors because of their ability to adapt themselves 
to human dwellings and because they use the human blood as 
source of feed. 
In Brazil, where the transmitters (or vectors) are Phebotominae 
insects in which the main transmitter genus is Lutzomyia spp., 
the transmission profile has been modified along the last 
decades, and the illness is now considered antropozoonotic 
(Tolezano, 1994; Gomes, 1992). At Säo Paulo, a Brazilian state, 
the etiological agent for CL is Leishmania braziliensis 
braziliensis species (Gomes and Galati 1989). In the study area 
located in Itapira municipality, in the inlands of Säo Paulo state, 
the probable CL transmitters are Lutzomyia intermedia, which 
have strong tendency to invade human dwelling, and Lu. 
whitmani, which is essentially sylvan (Gomes er al. 1989; 
Forattini, 1973). According to previous studies, the 
Phlebotominae flight range can vary from 200 to 1000 meters, 
but in most of the cases the range stays between 200 and 300 
499 
meters (Forattini 1973; Gomes and Galati 1989; Dourado ef al. 
1989: Gomes et al. 1989; Corte et al. 1996). 
In most of the cases the CL incidence, its prevalence and 
distribution, are influenced by the human activities related to 
deforestation, which propitiates the contact between people and 
environment, more specifically in the neighborhood of 
vegetation remnants (Forattini 1973; Dourado er al. 1989; 
Gomes et al. 1990; Gomes 1994). The Remote Sensing allows 
the identification of the vegetation cover and amount, along the 
time. The Geoprocessing techniques allows the user to make 
maps and quantify the vegetation involved in the life cycle of 
the CL transmitters mosquitoes (Hugh-Jonesl3 1989, 
Verhasselt18 1993, Washino and Wood 1994, Clarke et al. 
1996, Croner et al. 1996, Cross et al. 1996, Mbarkil et al. 
1995, Miranda et al. 1996, Hassan et al. 1999, Beck et al. 2000, 
Costa 2001). 
The deforestation is increasing all over the world requiring 
large scale analyzes, which can be provided by satellite images 
interpretation. If the vegetation related to CL transmission were 
identified and the deforestation patterns were discovered it will 
be possible to understand the relationship between human and 
vector populations, enabling a correct public health action. 
Thus, the objective of this study is to use Remote Sensing and 
Geoprocessing techniques to analyze the influences of the 
vegetation density and the deforestation pattern on the presence 
of CL transmitters in the neighborhood of the human dwellings, 
using 27 CL cases occurred between 1992 and 1997 in Itapira 
municipality.