Satellite Imaging and Surveillance of Infectious Diseases

ISSN: 2329-891X JTD, an open access journal Prevalence of Tropical Diseases J Trop Dis Satellite Imaging and Surveillance of Infectious Diseases Rajeev Singh1*, Koushlesh Ranjan2 and Harshit Verma1 1Department of Veterinary Microbiology, Sardar Vallabhbhai Patel University of Agriculture and Technology, Meerut, India 2Department of Veterinary Physiology & Biochemistry, College of Veterinary & Animal Sciences, Sardar Vallabhbhai Patel University of Agriculture &Technology, Meerut-250 110, India


Introduction
The sufficient amount of knowledge has accumulated over the past decades on relationship of the environment and disease, but present time demands the identification of the most robust environmental correlates of animal disease and accurately linking it with remote sensing technology in order to develop early and effective disease warning system. Recently, scientists have made efforts in using remote sensing technology in predicting the few vector-and waterborne diseases of humans and animals and it is expected that the role of remote sensing technology in developing emergency programs for disease surveillance, treatment and control will increase in near future. A brief review on the applications of satellite imaging in disease transmission, control and prevention is presented as under.

Infectious Diseases
The first effort of linking the satellite based environmental data with the disease was made in an eradication campaign of dracunculiasis at the Republic of Benin. Lateron, the satellite imaging was used to develop risk maps and control of few water and vector borne diseases.

Waterborne diseases
Water and climate are very closely linked. The excessive precipitation results in flooding which is a major cause of water-borne diseases viz. cholera, diarrhoea, typhoid, leptospirosis, hepatitis and even tetanus. The remote sensing technology can help the officials to determine the time and cause of contamination of water bodies used as source of drinking water and allows them to institute the rigorous prophylaxis and control measures in advance before the actual outbreak occurs. However, there are some requirements that must be met before predicting the effect of environmental changes on waterborne diseases viz., detailed knowledge of the disease incidence and transmission cycle, factors determining the contamination and survival of pathogen in water, specific and detectable indicators of microbial contamination of water and better health surveillance data and this information can be better used as determinants of waterborne disease outbreaks in remote sensing technologies.The prediction of a few of these waterborne diseases has been made using satellite imaging and is discussed as under.

Cholera:
The cholera is a bacterial disease of small intestine. It can be fatal to humans. The disease is caused by Vibrio cholera, a gram negative, non-spore forming and toxin producing curved organism. The infection may result in extreme diarrhoea, vomition, loss of ions and dehydration. The victims can die within a day or so unless ions and water are replenished quickly. The transmission occurs through faecal contamination of food and water from an infected person. The spread of disease can be checked through consumption of chemically disinfected or boiled water. The seventh cholera pandemic began in 1961, but, still has its effects in six continents. An estimated 3-5 million cases and over 10,000 deaths occur globally every year. The Vibrio cholera survives on the phytoplankton and zooplankton in inter epidemic period. The inorganic and organic content of water, pH, salinity, surface temperature and exposure to ultraviolet rays in sunlight affect the distribution of organism. The satellite is providing the new insight from space to track the blooms of tiny floating planktons that carry cholera bacteria, ocean temperatures, sea height and other related variables to predict outbreaks of cholera. In a study of this kind, the Advanced Very High Resolution Radiometer (AVHRR) sensed National Oceanographic and Atmospheric Administration (NOAA) satellite data was used to infer the presence and to monitor the temporal spread of cholera in order to develop the early warning system of cholera outbreaks in Bangladesh [1]. In the study, the public domain remote sensing data of 1992-1995, for Bay of Bengal were compared with the cholera cases in Bangladesh. The remote sensing data of sea surface temperature and sea surface height at the coastal region of Bangladesh were included in the study. The sea surface temperature was related to phytoplankton and zooplankton concentration and sea surface height was related to human plankton contact, as Bangladesh is only slightly above of sea level with tidal intrusion of plankton into inland water of Dhaka, where, water is consumed from river system without treatment. The superimposition of the data revealed a statistically significant correlation of the sea surface temperature, sea surface height and the annual cycle of cholera outbreaks. The extensive studies conducted during the previous decades, confirming the autochthonous existence of V. cholerae in the aquatic environment as commensal of zooplankton, i.e., copepods and present satellite data provides strong evidence of climate dependency of cholera outbreaks. This Bangladesh model of cholera prediction can be extended to the global scale as an early warning system for effective prevention of cholera outbreaks in cholera-endemic regions.

Colibacillosis:
The colibacillosis is a bacterial disease of human and livestock caused by E. coli. The E. coli is normal inhabitant of intestine, but the E coli that harbour virulence gene coding for colonization factors or toxins, produce disease. The diarrhoea, colic, vomition, septicaemia, dehydration, urinary tract infection and renal failure are major clinical manifestations. The coliform bacteria generally feed on decaying animal and plant tissue, whereas, E. coli feed on faecal matter. The faecal contamination of water is one of the main ways of disease transmission and the presence of E. coli in water is a strong indication of recent sewage or animal waste contamination. The satellite imaging may be helpful in monitoring of water quality and controlling future outbreaks in human population.
The LANDSAT-TM algorithms was developed for mapping the surface freshwater contents of cyanobacteria, total coliform bacteria and faecal coliform E. coli, the three main bacterial contaminants of drinking water sources [2]. The algorithm for cyanobacteria was developed based on the two pigments, chlorophyll a and phycocyanin, of which phycocyanin was targeted especially as it is the pigment found only in cyanobacteria. The phycocyanin content was measured in targeted surface water samples and algorithm was developed for mapping of bacteria in streams with widths ≥ 90m and drinking water reservoirs ≥ 5 hectares in area, for water depths of ≥ 2m. The same methodology was used to develop algorithms for total coliform bacterial (TC) and faecal E. coli (EC) contents in Lake and river water of ≥2 m depth. The results revealed that the LANDSAT-TM algorithm has a potential to play unique role in monitoring the water quality of small drinking reservoirs in cities, towns and villages throughout the world.
Later on, Low-Earth-Orbit Satellites was used in data communication networking between portable hydro-meteorological measuring stations that allowed automated event sampling and short time monitoring on increments of E. coli in the alpine Karstic spring water, the important drinking water resource for humans [3]. The activities of the eventsampling were monitored on Linux-Server based internet platform. The results of samples collected conventionally by hand and with the autosampling procedure were analysed and were found in agreement with ISO 9308-1 reference method. The two large summer events of 2005 & 2006 were monitored for comprehensive hydrological characterizations and detailed analysis of E. coli dynamics (n = 271) in spring water. The E. coli concentrations were corrected individually for event specific die-off rates (0.10-0.14 day −1 ) of E. coli to compensate the losses in the water samples stored at spring temperature in auto sampler. The high resolution temporal analysis of faecal contamination of drinking water revealed a sudden increase in the E. coli concentrations (approx. 2 log 10 units) in the spring water with a specific time lapse after the beginning of the event. The statistical analysis revealed that the spectral absorbent coefficient measured at 254nm may act as an early warning indicator for monitoring the time of faecal input in the drinking water.
Recently, a five year research project has been submitted, entitled Globolakes to monitor the water health of over 1000 lakes around the globe and to develop the first satellite-based early warning system [4]. This project was intended to study the effect of impact of climate change and associated factors on the algal blooms that can deplete the oxygen concentrations and produce the toxins harmful to human health in lakes and water reservoirs frequently used to supply the human drinking water. The Globolakes programme was builds on the satellites, Sentinel 2 and 3 of the European Space Agency's program, Global Monitoring for Environment and Security that will provide the images of the lakes.

Arthopod vector borne diseases
The vectors are the insect carriers' viz. mosquitoes, ticks, lice, fleas and flies etc that spread the germs to the humans or animals. These vectors transmit a number of deadly diseases viz. malaria, dengu fever, yellow fever, equine encephalitis, Japanese encephalitis, west nile fever, rift valley fever, louping ill, three day sickness, blue tongue, Q fever, ehrlichiosis, endemic thypus, bubonic plague, anaplasmosis, babesiosis, theleriosis, trypanosomiosis, leishmaniosis, tularemia, vesicular stomatitis, lyme disease etc. The environmental conditions such as rainfall and temperature, influence the vector biology along with the host biology as well as host-vector interaction and therefore to a greater extent the spread of vector borne diseases. The satellite measurements and other remote sensing techniques directly cannot identify the vector themselves, therefore, the satellite derived environmental variables viz. temperature, humidity, precipitation, land cover type, health of vegetation that are conducive for breeding of vector are used to identify and characterized the habitat in which vector thrives and these help in anticipating the outbreak of vector borne diseases.

Mosquitoes borne diseases:
The mosquitoes play an important role in existence, persistence, multiplication, development and transmission of a number of infectious agents that cause diseases and death in human and animals. A number of mosquitoes including Aedes, Culex, Anopheles, Psorophora, Mansonia spp are major parasites as well as vectors of infections in domestic animals. These vectors prefer still water for breeding. They lay their eggs in stagnant water of dams, ponds, shallow pits near the roadside, paddy fields, water bodies, standing water in channels of cultivated land, fountain water, tree holes, wet underside surface of vegetation, leaf axils, open drainage channels, water spilled near wells, man-made water containers such as domestic water storage containers, metal drums, automobile tyres, recycling containers, cooler water, swamps, marshy area and foot pits of animals. Most eggs hatch out in larvae in 48 hrs. The larvae feed in water and changes into pupa. The pupa develops into adult mosquitoes. The factors like temperature, rainfall, humidity, wind, direction, duration and speed influence the hatching and life cycle of mosquitoes. The seasonal density of mosquitoes in space and time, microbial adaptability in mosquito's biological life, abundance of infected mosquitoes, biting rates, inoculums per bite, proximity to breeding grounds and interaction with definitive host determine the ecology, demography, incidence and spread of mosquitoborne diseases. There is a long history of developing effective models for prediction and assessment of mosquitoes borne diseases based on environmental indicators and few are given as under. affects humans and animals. Amongst animals, cattle, sheep and goats are principle victim. Besides, the disease also occurs in wild ruminants and buffaloes. In humans, the disease is characterized by flu like fever, muscle pain, joint pain, headache, photosensitivity, vomiting and death in severely affected patients. In animals, infection results in fever, diarrhoea, abortion in pregnant animals and high mortality rate of new borns. The disease was first time reported in Kenya and there after remains confined in Africa. In 2001, the disease was reported for the first time outside Africa in Arabian Peninsula and increases the fear of expansion of area especially in direction of the Asia and Europe. The virus is primarily transmitted by mosquitoes of different species. In endemic areas, virus regularly circulates between ruminants and haematophagous mosquitoes. Certain Aedes species act as reservoir of virus during inter epidemic period and may transmit virus to their eggs. Increased precipitation in dry areas leads to an explosive hatching of mosquito eggs, many of which harbour RVF virus. Satellite imaging has been used to confirm the historic importance of precipitation in RVF outbreaks and in forecasting high risk areas in future.
The ground rainfall and moisture pattern data recorded from 1981 to 1988 through National Oceanic and Atmospheric Administration (NOAA) satellite 7 and 9 was used in measuring the green leafy vegetation dynamics as indicator of mosquito breeding habitat in Kenya and correlated it with the Rift Fever Virus ecology and used it as alarm for start of mosquitoes control prograame much earlier to actual commencement of RVF outbreaks [5]. The efforts were made in modelling the mosquitoes breeding habitat in East Africa. The mosquito breeding habitats in East Africa are locally known as "dambos". The rain water gets accumulated in "dambos". The still water in "dambos" buildsup the vector mosquito population. The high density of mosquitoes increases the RVF outbreaks in susceptible hosts. In the model testing the satellite derived normalized difference vegetation index (NDVI) was correlated with the cause input of variations in rainfall and effect output of fluctuation in RVF incidence. Further, the global precipitation impact of El Nino Southern Oscillation and concurrent elevations in sea surface temperature (SST) in the Pacific and Indian oceans when incorporated as associated variables along with increased rainfall in Eastern Africa, improve the prediction of RVF up to 5 months in advance [6,7].
Landset data was used in determining the green leaf area index (LAI) of rice fields and relating it with the build-up of Aedes mosquito density [8]. They compared LAI with number of Aedes freeborni larval present at the outskirt of the rice fields and the minimum distance between the centre of each field and the nearest animal pastures serving as source of blood meal. The results revealed that rice fields with high LAI and near to animal pastures produced more numbers of mosquitoes, compared to fields with low LAI and far from the pastures. The discriminate analysis used the spectral measurements obtained from the satellite and minimum distance to animal pastures to accurately identify the high mosquito producing areas.

West Nile virus:
West Nile is mosquito borne viral disease that causes encephalitis and fatal neurological disorders in human and animals. The disease has been reported from Africa, Asia, Europe and United States. The virus is maintained in nature in a cycle involving transmission between wild birds and mosquitoes. Wild birds act as amplifying host. Humans, horses and other mammals can be infected through bite of infected mosquitoes and act as dead end host. The virulent strain also causes encephalitis and paralysis in crows. The Culex mosquitoes are generally considered the principle vector and in low rate virus is maintained in mosquito's population through vertical transmission.

The migratory birds introduce the virus in new areas.
Climate variability is one of the most important factors influencing the virus multiplication and circulation; vector abundance, biology, physiology and vector contact to susceptible avian communities and may help in forecasting the future WNV disease risk areas. The everyday reporting of dead birds, active virus surveillance of birds, reports of infection in humans and horses, virus detection in mosquito's vector along with satellite derived virus spread sensitive climatic variables may help in control or exclusion of disease in risk identified areas [9].
The multiple environmental predictors viz. land surface temperature (LST), normalized difference vegetation index (NDVI) and actual evapotranspiration (ETa) data derived from moderate resolution imaging spectroradiometer (MODIS) to was used to develop predictive model of West Nile virus transmission risk in humans of northern great plains of United State, which is considered as hotspot of human disease and found that environmental monitoring using remote sensed data have a good space in surveillance of West Nile virus risk prediction in space and time [10].

Japanese encephalitis:
The Japanese encephalitis (JE) is another mosquito-borne viral disease. The disease is caused by Japanese encephalitis (JEV) virus of genus Flavivirus. The wild birds (especially herons) and domestic pigs act as reservoirs for JEV and transmission to human host may lead to severe symptoms. It is endemic in several parts of the world including India and Republic of Korea. The disease is caused by Japanese encephalitis (JEV) virus of genus Flavivirus. The mosquito Culex tritaeniorhynchus transmit the disease to humans. The experts note an increase in the density of Culex tritaeniorhynchus with the cultivation of rice in field. The rice fields provide site for breeding and shelter required for vector Culex tritaeniorhynchus. The rice field satellite surveillance showed a linear relationship between area of rice field cover and mosquito trapped in that area [11]. The Verma and Gupta (2013) successfully used the LANDSAT ETM remote sensing data of rice field as potent ground for mosquito vector breeding place and related them with the JE outbreaks in Gorakhpur district of Uttar Pradesh state [12]. The other closely related mosquito-borne Murray Valley encephalitis virus is enzootic in northern Australia and Papua New Guinea. The disease is directly related to tropical rain fall in these areas. The Schuster et al. (2011) used the remote sensing (RS) data of Tropical Rainfall Measurement Mission (TRMM) in developing the logistic regression models based on Multi-satellite Precipitation Analysis (TMPA) and recorded a direct relation between tropical rain fall and MVEV outbreak in these areas [13]. The mosquitoes breed in standing water bodies that can present for more than two weeks to support the aquatic phase of life cycle. The too little rains create fewer breeding habitats. The heavy rains destroy and wash away the larvae and eggs. The extremes of temperature also affect the life cycle. The mosquitoes develop at optimum temperature of 27 °C. The development of the plasmodium parasites within the mosquito vectors also depends upon the temperature. The Plasmodium falciparum sporozoites complete its development in gut and to reach to the salivary glands of mosquitoes in about 10 days. It can be more or less depending upon the decrease or increase in temperature. A variation in extrinsic incubation period in parasite in mosquito's vector have been recorded with fluctuation of air temperature from optimum. The disease is seasonal and incidence of malaria correlates well with temperature, rainfall and humidity patterns. However, irrigated paddy fields, marshes and wetlands are very favourable for the breeding of mosquito species.
The environmental variables that determine the habitat climatology and subsequent build-up of mosquito population density along with infectious bite and quantum of parasite in inoculums have been used to model the malaria. Ford (2012) used the amount and location of stagnant water supporting the mosquitoes breeding in forecasting the malaria outbreak through satellite imaging [16]. The Rogers et al (2002) discussed in detail the use of satellite images in describing the distribution of important species of Anopheles gambiae complex and the number of infectious bites per person per year (entomological inoculation rate/EIR) to predict the risk of malaria in Africa in a discriminate analytical model [17]. The Charoenpanyanet and Chen (2008) used Landsat 5 TM data of land cover and related it with the densities of Anopheles mosquitoes vector to develop predictive model of malaria in western Thailand [18]. Rahman (2011) developed a predictive model for malaria vector distribution in Bangladesh from meterological data collected from remote sensing [19].

Culicoides borne diseases
Bluetongue: The bluetongue (BT) is a vector borne (Culicoides) disease caused by 27 different serotype of bluetongue virus (BTV). The BTV outbreaks can be forecasted by prediction of Culicoides migration in a geographical area. The BT forecasting model based on presence of C. imicola vector across Europe and north Africa was validated which predicted the presence and high abundance of BT epizootics, in several regions including Sardinia, Balearics, Sicily, Corsica, areas of mainland Italy, large areas of western Turkey, Greece and northern Algeria and Tunisia [20]. The Guis et al., (2007) used the high resolution satellite imagery to identify the environmental variables assisting in BT outbreak in several area of Europe including Corsica, a French Mediterranean island [21]. The Hartemink et al., (2009) used the basic reproduction number (R₀) of BTV to estimate the risk of BT outbreak in an area after its introduction [22]. In Netherlands, integrating the satellite imagery data with vector abundance and biologically mechanistic modelling, BTV outbreak was predicted. The NDVI ranks first and is closely followed by CCD and land surface temperature amongst the input variables used for monitoring the tsetse flies distribution in the model. Roger and William (1993) reviewed the different methods used in modelling arthropod vectors from satellite data [23].

Ticks:
The metrological data from environmental satellite are of great value in selecting the input indicators for modeling the effect of climate on the life cycle and epidemiology of tick borne diseases. The climate changes can explain about fifty five per cent of tick-borne encephalitis. A positive association between preceding people's visit to forests and succeeding increase in the number of cases of tick-borne encephalitis was recorded. In this way the satellite monitoring of climate change and human population movement can help in predicting and preventing the outbreaks of this vector-borne disease.

Crimean-Congo haemorrhagic fever:
The Crimean-Congo haemorrhagic fever (CCHF) is a widespread disease caused by a tickborne Nairovirus of Bunyaviridae family. It is endemic in several parts of the world including Africa, Middle East and Asia. The satellite based surveillance of ground vegetation, agro climatic condition and tick prevalence in area can be correlated with disease outbreak. In Turkey, a predictive model using Landsat images system of satellite surveillance was developed to map the habitat suitability for tick vector. The area having higher report of CCHF were directly correlated (p<0.05) with area of higher climatic suitability for tick vector [24]. The mediumresolution MODIS satellite imagery showed the potential spread and mixing of H. marginatum tick populations from eastern Europe viz. Turkey, Russia, and Balkans to western Europe viz., Italy, Spain and northern Africa [25]. This type changes might be possible due to climatic changes which replicates the virus speared in newer areas. Debien et al., in a satellite based study established the relationship between the precipitation and greenness and observed a positive correlation between the precipitation and one or two months lagging higher normalized difference vegetation index (NDVI) [26]. Finally they found the positive association in the cases of plague and climatic variables.

Hanta virus infection:
Hanta virus pulmonary syndrome is a serious illness of humans. The symptoms are flu like and include fever, chill, headache, nausea, vomiting, muscle ache, diarrhea, abdominal pain, coughing and shortness of breath. The mice and rats spread the disease. They excrete the virus in the urine, droppings and saliva which form the tiny air droplets of virus. The breathing the infected air or contact of infected rodents or their urine or droppings transmits the disease in the people. There is no specific chemotherapy. The vaccine is also not available. The control of rodents in and around of house is best way to prevent the infection.
Studies have been conducted to predict the rare but deadly Hantavirus outbreaks in months ahead of time using satellite images through monitoring surges in vegetation that boosts mouse population. The numerous satellite images were used to understand the variations in the amount of green vegetations covering the Earth's surface and to track mice, the carrier of Hanta virus. It was observed that rise in vegetation create more food for the mice with increase in mice population, contact rate of mice and human that leads to more outbreaks of Hanta virus in humans. and reported that increased precipitation succeeding the prior drought years in northern New Mexico and southern Colorado increased the risk for HPS [27]. Later on, in 2011 Cao et al., combined the satellite imagery along with total number of captured mice, proportion of infected mice with the disease outbreaks and found that both climbed after peaks in greenery [28].

Bat vector borne diseases
Rabies: Rabies is a highly fatal viral disease of serious concern in different parts of world including Asia and America. The satellite telemetery surveillance and GIS data were used for study the disease at coarse scale. The bat (Tadarida brasiliensis) borne rabies in Chile during 2002-2012 was studied using ecological niche model (ENM) to find the association of rabies occurrence and environmental factors. The rabies virus variant AgV4 in bat was related with the Normalized Difference Vegetation Index obtained from Moderate Resolution Imaging Spectroradiometer satellite. The scenario build for rabies in bats and reliable anticipation of human rabies revealed the usefulness of ENM for rabies and other zoonotic pathogens [29]. The Satellite surveillance was also helpful in dispensing of vaccine products to its target wild animals. The Bait vaccines against rabies were distributed aerially in a region having large number of wild foxes (Vulpes vulpes) using a satellite navigated automatic bait drop system in Italy [30].

Wild birds borne diseases
Avian Influenza: Satellite surveillance of waterfowl on tracks of migratory birds may help in understanding the transmission cycle of highly pathogenic avian influenza. For this purpose the satellitederived movement of eight species of ducks captured in water bodies at Hong Kong, Bangladesh and Turkey were analysed [31]. Satellite based surveillance of migratory birds and Avian influenza specially in South east Asia was done by several researcher [32,33]. Satellite telemetry based surveillance was used for Brown-headed gulls (Larus brunnicephalus) in winter season in Thailand during 2008-2011 for investigating their roles in spread of highly pathogenic avian influenza H5N1 virus. However, the surveillance did not show any bird infected with H5N1 virus during the study period and no H5N1 outbreaks were reported in Brown-headed gulls during the test period [34].

Conclusions
The gradual change in climate due to global warming are bound to alter the ecology, habits and habitats of hosts, agent and vectors and so drastically that incidences of vector and waterborne diseases such as malaria, dengu, diarrhea are bound to increase in endemic areas. Further, these diseases may expand the geographical range and new epidemic and pandemic of these diseases may occur in areas previously free of disease. So there is need of (1) construction of potential risk areas maps based on conditions favorable to vector proliferation of exotic diseases before their entry in new country, (2) development of early warning system before harmful exposures, (3) improved prevention initiative targeting ingress of risk, (4) reduction of environmental-related diseases, and (5) improvement in bio-terrorism event information management. The use of satellite in prediction and monitoring of diseases is a very recent & exciting approach, though in infancy and needs more refinement yet has potential of worth due to high speed in time series.