Skip to main content
Download PDF

Examining noncommunicable diseases using satellite imagery: a systematic literature review

BMC Public Health volume 24, Article number: 2774 (2024) Cite this article

Abstract

Introduction

Noncommunicable diseases (NCDs) are the leading cause of morbidity and mortality worldwide, accounting for 74% of deaths annually. Satellite imagery provides previously unattainable data about factors related to NCDs that overcome limitations of traditional, non-satellite-derived environmental data, such as subjectivity and requirements of a smaller geographic area of focus. This systematic literature review determined how satellite imagery has been used to address the top NCDs in the world, including cardiovascular diseases, cancers, chronic respiratory diseases, and diabetes.

Methods

A literature search was performed using PubMed (including MEDLINE), CINAHL, Web of Science, Science Direct, Green FILE, and Engineering Village for articles published through June 6, 2023. Quantitative, qualitative, and mixed-methods peer-reviewed studies about satellite imagery in the context of the top NCDs (cancer, cardiovascular disease, chronic respiratory disease, and diabetes) were included. Articles were assessed for quality using the criteria from the Oxford Centre for Evidence-Based Medicine.

Results

A total of 43 studies were included, including 5 prospective comparative cohort trials, 22 retrospective cohort studies, and 16 cross-sectional studies. Country economies of the included studies were 72% high-income, 16% upper-middle-income, 9% lower-middle-income, and 0% low-income. One study was global. 93% of the studies found an association between the satellite data and NCD outcome(s). A variety of methods were used to extract satellite data, with the main methods being using publicly available algorithms (79.1%), preprocessing techniques (34.9%), external resource tools (30.2%) and publicly available models (13.9%). All four NCD types examined appeared in at least 20% of the studies.

Conclusion

Researchers have demonstrated they can successfully use satellite imagery data to investigate the world’s top NCDs. However, given the rapid increase in satellite technology and artificial intelligence, much of satellite imagery used to address NCDs remains largely untapped. In particular, with most existing studies focusing on high-income countries, future research should use satellite data, to overcome limitations of traditional data, from lower-income countries which have a greater burden of morbidity and mortality from NCDs. Furthermore, creating and refining effective methods to extract and process satellite data may facilitate satellite data’s use among scientists studying NCDs worldwide.

Peer Review reports

Introduction

Noncommunicable diseases (NCDs) account for 74% of global deaths annually, with cardiovascular diseases, cancers, chronic respiratory diseases, and diabetes responsible for over 80% of premature NCD mortalities [1]. NCDs are not limited to older adults, with 17 million deaths before age 70, predominantly in low- and middle-income countries [1]. In the United States, direct health costs related to NCDs exceed $1 trillion annually [2]. Unhealthy behaviors like smoking, poor diet, and physical inactivity increase NCD susceptibility [3]. The third largest underlying risk factor of chronic disease (after high blood pressure and tobacco usage) is air pollution, an environmental risk factor most often increasing the risk for three of the top four NCDs - cardiovascular disease, cancers, and chronic respiratory diseases [4].

Achieving the World Health Organization Sustainable Development Goal Target 3.4 of reducing premature NCD mortality by one-third by 2030 [5] is challenging, with most countries making minimal progress [6]. Identifying geographic locations with populations most at risk for NCDs is one step toward directing prevention-related policies and programs to achieve this goal [7]. Satellite technologies offer tools to help identify at-risk geographic locations that overcome limitations of traditional, non-satellite-derived environmental data (e.g., surveys and ground monitoring stations) such as subjectivity and being limited to smaller geographic areas. Satellite data is open source, available on a global scale, and has four resolutions: temporal, spatial, radiometric and spectral [8]. With over 400 Earth observation satellites orbiting our planet [9], satellite imagery data, often coupled with artificial intelligence (AI), has shown great promise in advancing areas of research outside of healthcare, such as earth science [10] and economics [11], and within health sciences, particularly in infectious disease [12,13,14]. While satellite data helps mitigate the problem of traditional environmental data availability, it presents new challenges in understanding what satellite data to use and how to interpret the data. Fortunately, publicly available algorithms, tools, and tutorials exist to help scientists extract, process, and interpret satellite data. Satellite imagery has been less commonly used for analyzing and managing NCDs [15].

Traditional (non-satellite-derived) environmental measurements have been successfully used in research in the form of surveys (e.g., light at night (LAN), greenspace) and ground monitoring stations (e.g., air pollution), both of which have limitations [8]. Survey data can be subjective and non-uniform, while ground monitoring station data is limited to areas within a close proximity to a station, most often a developed urban area [8]. Satellite data has been found to overcome these limitations by its being open source, available on a global scale, and having four resolutions: temporal, spatial, radiometric and spectral [8].

Satellite data is derived from remote sensors located on satellites. The amount of energy reflected, absorbed, or transmitted by any item on Earth creates a “spectral fingerprint.” Remote sensors can detect a number (specific to the type of remote sensor and called its spectral resolution) of spectral bands, which allows items to be identified by their spectral fingerprint [16, 17]. There are two types of sensors: passive sensors (e.g., radiometers and spectrometers operating in the visible, infrared, thermal infrared and microwave electromagnetic spectrum) that measure land and sea physical attributes (e.g., temperature, vegetation properties, aerosol properties, cloud properties) and active sensors (e.g., radar sensors, altimeters operating in the microwave band of the electromagnetic spectrum) that measure vertical profiles of land and sea attributes (e.g., forest structure, ice, aerosols). Satellites have specific orbits and sensor designs that dictate resolution [16]. How well a remote sensor can distinguish between small differences in energy is called its radiometric resolution, which is the amount of information in each pixel (e.g., 8 bit resolution that can store up to 256 values). Higher resolution means more detail, though this also requires more processing power. Spatial resolution is defined as the size of each pixel. For example, to see buildings you would need 10 m (m) spatial resolution, which represents a 10 m by 10 m square on the ground. Neighborhoods need 20 m spatial resolution, which represents a 20 m by 20 m square on the ground, while regional needs 1 km (km), which represents 1 km by 1 km square on the ground (national: 10 m, continent: 30 km and global: 110 km) [16]. Spectral resolution is defined by both the number of bands and how narrow the bands are. For example, 3–10 bands is referred to as multispectral, whereas hundreds or thousands of bans are hyperspectral. Temporal resolution is defined as the time it takes the satellite to complete one iteration of its orbit, which is dependent on its orbit, its swath, width and the specific sensor (e.g., Moderate Resolution Imaging Spectroradiometer (MODIS) on NASA’s Terra and Aqua satellite’s temporal resolution is 1–2 days) [16].

We aimed to provide the first systematic literature review focusing on how environmental factors data collected from satellite imagery has been used to examine risk, incidence, prevalence, or mortality related to an NCD, both methodologically and technically. This review can illuminate resources and methods for using emerging satellite imagery technologies to capture and analyze comprehensive data that can inform NCD prevention and control interventions and policies. By integrating satellite-derived data with ground-based monitoring systems, scientists and policymakers can better understand the risk and distribution of NCDs, allocate resources more effectively, and implement targeted strategies to lessen NCD burden.

Methods

This systematic review was registered at PROSPERO (CRD42023433472). We followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guidelines [18].

Data sources

We conducted a systematic review of literature related to satellite imagery and the top 4 NCDs (cardiovascular diseases, cancers, chronic respiratory diseases, and diabetes) in the world through June 6, 2023. We did not restrict our search to any start date. To gather relevant studies, we searched PubMed (including MEDLINE), CINAHL, Web of Science, Science Direct, Green FILE, and Engineering Village databases. See Additional file 1 for keyword search strings.

Study selection

First, we removed duplicate studies. Next, at least two study authors independently assessed the remaining abstracts based on predetermined inclusion criteria of needing to examine the top four NCDs in the world using satellite imagery. We considered all quantitative, qualitative, and mixed method study designs written in English. Then 2 study authors independently evaluated the full-text articles for inclusion, with discrepancies resolved through discussion. Studies that were not about one or more NCDs, not about satellite imagery, or review articles were excluded. Another way to describe our inclusion criteria is using the PECO (Population, Exposure, Comparator, Outcomes) framework recommended for exploring associations of environmental and other exposures with health outcomes [19]. Table 1 presents the inclusion and exclusion criteria.

Table 1 Inclusion and exclusion criteria
Full size table

Data extraction

At least two study authors independently extracted information for each study that met the inclusion criteria, including the study aim, disease, geographic level, year of data collection, methods, tools and resources, data extracted from images, measures, results, and findings. The authors discussed and resolved any discrepancies in the extracted data. We assessed the quality of the evidence for each study using the criteria from the Oxford Centre for Evidence-Based Medicine [20]. The quality of each study was independently graded by two study authors, with any discrepancies resolved through discussion. Following is a description of the quality ratings: 1 for properly powered randomized clinical trials, 2 for well-designed controlled trials without randomization and prospective comparative cohort trials, 3 for case-control studies and retrospective cohort studies, 4 for case series with or without intervention and cross-sectional studies, and 5 for case reports or opinions of respected authorities.

We conducted a qualitative synthesis of satellite data by determining if an association (statistically significant relationship) was found between the satellite data and each study’s dependent variable (e.g., NCD outcome) and explored the authors’ statements about the value of using satellite data. We also recorded statements that included wording about satellite data such as “overcame the problem,” “great tool,” and “enhanced.” We used a spreadsheet so that at least two authors could track these associations and statements and used codes to categorize aspects about the value of satellite imagery for examining NCD outcomes for each article. Our analysis also included all authors reviewing the frequency of the findings and the wording of the statements. Based on this analysis, the authors developed themes about the value of satellite imagery for examining NCDs.

Results

We identified 1,495 articles from our database searches. After applying inclusion and exclusion criteria, 43 studies were selected for inclusion and 1,452 were excluded (Fig. 1). Table 2 includes details, quality assessment, and study authors’ statements about satellite value for all the reviewed studies and Table 3 presents the study characteristics, analysis, and data synthesis.

Fig. 1
figure 1

PRISMA flow diagram

Full size image
Table 2 Study details, quality, and authors’ statements about satellite value in reviewed articles
Full size table
Table 3 Characteristics of the studies, analysis, and data synthesis (n = 43)
Full size table

The study publication dates spanned over 15 years, 2008–2023, with more than half published within the 5 years before our study search end date. Overall, 70% of the studies were from high-income countries, with over half of those from the United States. The remaining studies were from middle-income countries, and none were from low-income countries. The majority of studies (66%) used satellite data examined at the city, census tract or census block, or county level. About half of the study designs were retrospective cohort and about one-third were cross-sectional. The 12% of studies that fit in the most rigorous study design category for this review (had prospective study designs) all focused on cancer. There were no randomized control trials or case reports. Regarding disease outcomes examined, prevalence was an outcome in over half of the studies (53.5%), incidence in about 30% of the studies, mortality in just under 20%, and disease risk in just under 5% (Table 2).

30% of the studies used satellite images from MODIS [21,22,23,24,25,26,27,28,29,30,31,32,33,34], an instrument located on NASA’s Aura, Terra, and Aqua satellites, and 16% used unspecified instruments located on NASA’s Landsat satellites [24, 35,36,37,38,39,40,41,42,43,44,45]. The other major satellite image source was the United States Space Force’s Department of Meteorological Satellite Program (DMSP) [46,47,48,49,50,51,52], with images used in 21% of the studies. MODIS and Landsat were primarily used to extract air pollution and greenspace data, while DMSP was primarily used to detect LAN. Greenspace was the most frequently extracted data, with nearly half of the studies examining this feature. The next most frequently extracted data was air pollution, appearing in 37.2% of the studies (Table 3).

All 4 major NCDs examined—cardiovascular disease, cancers, chronic respiratory disease, and diabetes—appeared in at least 20% of the studies, with chronic respiratory diseases and cancers each appearing in about 40%. Cardiovascular disease, chronic respiratory disease, and diabetes studies heavily used greenspace and air pollution data. All studies using LAN examined cancer outcomes [46,47,48,49,50,51,52], with 4 of the 7 specific to breast cancer [46, 47, 50, 51]. DMSP data was from 1996 to 97 and was used as a baseline for determining LAN in 6 [47,48,49,50,51,52] of the 7 articles [46,47,48,49,50,51,52]. Approximately one-third of the cancer studies included air pollution data [23, 25, 29, 35, 53, 54], while less than one-fifth included greenspace data [27, 53, 55] (Table 3). Air pollution was the primary data extracted for chronic respiratory disease, closely followed by greenspace. We found the reverse with the studies on cardiovascular disease; greenspace was the primary data extracted, closely followed by air pollution. Greenspace was the primary data extracted for diabetes studies (Fig. 2).

Fig. 2
figure 2

The number of studies of each type of satellite data for each noncommunicable disease. “LAN” is “light at night” and refers to ambient light exposure at night. “Extracted features” refers to features extracted using machine learning to inform a machine learning model. “Air pollution” refers to aerosol optic depth measures, particle matter in air with diameter less than 2.5 micrometers, particle matter in air with diameter less than 10 micrometers, “Flood” refers to changes in land and water surface due to rainfall. “Greenspace” refers to normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), photosynthetic vegetation (PV), and soil adjusted vegetation index (SAVI). “Temperature” refers to the land surface temperature (LST) and surface urban heat island (SUHI)

Full size image

The majority of the articles (60%) used data from earlier than 5 years from the publication date and matched health data year(s) to satellite data year(s). Of the 10 articles [26, 37, 41, 45, 47,48,49,50,51,52] using data from earlier than 10 years from the publication date, 6 used data from DMSP [47,48,49,50,51,52], and 3 used data from LandSat [37, 41, 45]. All but 3 articles [39, 48, 49] in the review found associations between the satellite data and NCD outcomes of risk, incidence, prevalence, or mortality. For example, Fan and colleagues [38] found a positive association between neighborhood greenness and COPD prevalence using the NDVI based on satellite imagery. Two of the studies that found no association used LAN and cancer [48, 49]; however, both found that LAN was a valid representation of circadian rhythm disruption. The third article found no association between greenspace in early life and insulin resistance in adolescence [39] (Table 2).

All data extracted in the studies in this review were related to a previously known disease risk factor. That is, no studies introduced a new disease risk factor that had not been established in prior research. Additionally, the majority of the studies included covariates such as sociodemographic factors like age and income level or health-related factors like body mass index (BMI) and smoking. Most studies focused on one type of satellite data, such as greenspace. However, 9% of the studies [22, 24, 28, 37, 38, 40, 44, 53, 56] examined multiple environmental factors extracted from satellites, such as greenspace and air pollution, in studying the incidence [28, 44], prevalence [22, 24, 37, 38, 40, 56], and mortality [28, 53] of NCDs. Two articles used a type of AI, convolutional neural networks (CNN), to extract numerical features from satellite imaging. One article [57] used a CNN with t-SNE (t-distributed Stochastic Neighbor Embedding) to verify the capacity of the neural network to extract relevant features related to cancer prevalence. The other article [58] used a visual geometry group fast convolution neural network (VGG-CNN-F), a network previously described by Chatfield and colleagues [59] with elastic net regression to prevent overfitting of the model to training data and minimize mean cross-validation error in a study examining obesity prevalence.

All articles used satellite data extracted based on geolocation(s) specific to the population of interest’s location (versus using data from a convenience sample based on data availability) using one or two of a variety of methods (Table 3). Existing publicly available algorithms were the primary method for satellite data extraction, with such algorithms used for analyzing 85% of greenspace data (primarily the normalized difference vegetation index, or NDVI) and 50% of air pollution data (such as the multi-angle Implementation of Atmospheric Correction (MAIAC) and deriving aerosol optical depth (AOD) with geographically weighted regression (GWR), primarily for particulate matter 2.5, or PM2.5). Two articles used the Jenk’s Natural Break method algorithm to classify LAN data [46, 47]. Image preprocessing methods performed on raw satellite image data to prepare it for further data processing were referenced in just under half of the studies. Examples of image preprocessing are LAN data transformed into radiance [49, 50, 52] or determining image inclusion according to criteria such as 10% or less cloud cover. Image preprocessing calibration, gauging the data with a standard scale, was used in two articles [50, 60]. One article calibrated LAN with satellite sensor data to provide average daily radiance [50] and another calibrated satellite air pollution data with ground data using a land-use regression model [60]. One-quarter of the studies referenced external resource tools that extract data from satellite images, such as ArcGIS, MERRA-2, NASA’s Giovanni tool, and the Sentinel Application Platform [26, 28, 29, 46, 47, 51,52,53, 56]. Five articles used models, three of which were specific to chemistry: WRF-Chem [28], GEOS-chem [30], and an unspecified global chemical transport model [54]. Yuan and colleagues [54] also used the radiation transfer model and the differential absorption spectroscopy inversion technique. Qu and colleagues [31] used a spatial-statistical model with the NDVI to derive an estimate of residential greenness, and Qazi and colleagues [42] used a model for supervised classification. The “satellite data extraction method(s)” section in Table 3 shows the breakdown of satellite data extraction methods used in the studies.

The vast majority (90%) of studies found an association between their dependent variable and the satellite-derived data [21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38, 40, 42,43,44,45,46,47, 50,51,52,53, 55,56,57,58, 60,61,62,63,64,65] (Table 2). For example, both Bauer and colleagues [46] and Kloog and colleagues [47] found breast cancer incidence was associated with high LAN exposure. Fan and colleagues [38] found a significant positive association between COPD prevalence and greenness (NDVI). Another 2% of studies [39] found value in satellite data but no association with their dependent variable [39]. In this article, Jimenez and colleagues [39] found that NDVI is the most widely used satellite-derived indicator of green space and can be used as a longitudinal exposure measure. Furthermore, 58% of the articles stated that satellite data overcomes the difficulties present in research when geographic areas do not have environmental data or the available ground data is sparse [21,22,23,24,25,26, 28, 30, 32,33,34,35,36, 38, 39, 42, 44, 51, 53, 55, 57, 58, 63,64,65]. For example, Prud’homme [30] (2013) and Yitshak [64] (2015) specifically noted the limitation of ground data for air pollution, proximity to monitoring stations, and sparse spatial data, which can be overcome by using satellite data. The final two columns of Table 2 shows if the dependent variable and satellite-derived data were related and any claims stated by the study authors about the value of using satellite data. We found two themes regarding the value of using satellite imagery to examine risk, incidence, prevalence, and mortality related to NCDs. The first theme was that satellite data overcomes problems of sparse or missing spatial and temporal data. Traditional environmental data (non-satellite data) is limited by the range of each sensor and the completeness of data. Satellite data complements traditional data and extends the availability of environmental spatial and temporal data to a global scale. The following representative quote illustrates such sentiments:

In particular, we have clearly shown that, thanks to data availability and big data technologies, it is now possible to jointly study heterogenous data, such as health care and air pollution information extracted from satellites. This provides an unprecedented opportunity to improve our understanding of phenomena by extracting unseen temporal and spatial correlations [36].

Use of satellite images has become a great tool for epidemiology because with this technological advance we can determine the environment in which transmission occurs, the distribution of the disease and its evolution over time [44].

The second theme was the ability of open-source satellite data to enable studies to be extrapolated to other areas. Traditional data sources are specific to a distinct geographic area, whereas satellite data often allows for (spatial and temporal) data to be available globally. The following representative quote illustrates such sentiments:

The results of studies like ours could be extrapolated to other cities, given that the use of Sentinel 3 satellite images lies within the reach of the entire scientific community, and their use for determining land surface temperature (LST) and surface urban heat island (SUHI) is straight forward [53].

Our findings will have important public health implications for policy makers when they are planning the size, shape, density and accessibility of surrounding green spaces in living areas [38].

Discussion

Satellite imagery has been used in a variety of ways to address NCDs. A key finding of this review is that nearly all studies found an association between at least one type of satellite data and one NCD. Most of the studies used existing publicly available algorithms to extract data from images. Furthermore, a couple of studies [57, 58] attempted to harness the power of AI to see beyond predetermined features.

Over half of the studies analyzed satellite data based on geolocation(s) specific to the research population of interest’s location at the city, census tract, or census block level. One reason the authors may have chosen these geographic levels could be because sociodemographic and health-related data, types of data that have been integral to NCD studies [3], is often available at these levels [3, 66]. Furthermore, by adding satellite data at these geographic levels, study authors may have been able to fill gaps in environmental factors when using ground data alone, such as in Allen and colleagues’ [35] examination of air pollution in the city of Ulaanbaatar, Mongolia. Just one article used satellite data globally; it replicated previous studies that either excluded areas without ground monitoring or were limited by “coarse spatial resolution” [23].

Over 80% of premature deaths caused by NCD occur in lower-income countries [1], yet just one-quarter of the studies in our review were from lower-income countries. Satellites orbiting the Earth capture data from across the globe, including areas without robust healthcare data collection, such as lower-income countries [8]. Using satellite data for these countries could better enable disease surveillance, tracking health trends and risk factors, and informed healthcare decision-making [15, 67]. One potential strategy for using satellite data in lower-income countries is identifying trends between certain satellite-derived data and health outcomes in similar countries or regions that have health data available and extrapolating those trends to the lower-income country of focus. A similar method has been succeesfully used with satellite imagery data in the field of economics [68]. However, a limiting factor for using satellite imagery for health is a lack of proper tools, knowledge, and skills in collecting and analyzing satellite data among researchers, engineers, and government employees [69]. This is particularly problematic in lower-income countries with fewer educational resources to train people [69, 70]. One potential solution to address the lack of training is to use the publicly available governmental and university-based tutorial programs and resources designed to make satellite data easier to use [69,70,71,72]. Additional file 2 presents resources to help researchers, scientists, and policymakers understand, find, and use satellite imagery data.

Most reviewed articles discussed how satellite data is an asset to NCD research by providing open-access environmental data that surmounts the constraints of ground-based data collection methods and availability. However, it is important that investigators consider tradeoffs between levels of spatial, temporal, and spectral resolution when choosing their satellite remote sensing data sources [15]. Researchers can use guidance from organizations such as NASA [73, 74] and the European Space Agency [75, 76] to help determine the optimal scale of data needed for their research. For example, MODIS (on Terra, Aura, Aqua, and Sentinel 1a satellites) is an instrument that produces moderate-resolution images, while Sentinel 2 and 3 instruments produce high-resolution images. The choice between using MODIS or Sentinel 2 or Sentinel 3 would depend on the investigator team’s resources (e.g., computing processing power and available investigator hours) and the data needs (e.g., general greenspace in a city versus specific greenspace by city block) for their research questions.

While satellite imagery has existed for over a half-century [77], not until recently did scientists have a method to process the vast amounts of data amassed by observing our planet from space [78, 79]. AI can quickly and efficiently extract meaningful patterns, trends, and insights from satellite images. Most articles in this review (95%) examined previously known environmental factors (e.g., air pollution, LAN) using satellite image data. The paucity of studies employing AI to analyze satellite image data for NCD research–just two studies in this review–is notable given that other fields (e.g., waste management [80, 81], agriculture [82, 83], urban planning [82, 84], and defense [82, 85]) have more readily adopted AI methods to process satellite image data. Within health sciences, satellite imagery coupled with AI has been used to study patterns in infectious diseases [12, 13, 71]. With 241 in-orbit Earth observation satellites registered with the United Nations and that number growing [86], there is an opportunity to expand the use of AI and satellite data to monitor and analyze NCD risk for informing policy and programmatic decisions to improve noncommunicable disease outcomes.

Several studies used existing algorithms to measure satellite data. These algorithms allowed the researchers to avoid developing a new measure and better enabled comparisons between studies. We recommend that researchers continue testing these existing algorithms and develop and test new algorithms to measure satellite data to help facilitate future satellite imagery-based NCD research. Such algorithms could be constructed by researchers or generated using AI and validated through research studies. The existence of previously validated algorithms may help epidemiologists and other individuals focused on studying NCD conduct more robust satellite data-based studies and avoid the need to create and study the properties of a new measure.

Our study deviated from the original protocol in a few ways. First, due to the quantitative nature of the studies, we ended up using the assessment of quality from the Oxford Centre for Evidence-Based Medicine instead of the Mixed Methods Appraisal Tool. Second, we focused on the top four NCDs in the world because the World Health Organization highlights these four diseases as the deadliest diseases [1]. Third, we added more technical engineering and environmental databases to our literature search to ensure we captured as many articles as possible that fit our search criteria.

This review has some limitations. First, it was limited in scope to peer-reviewed literature; thus, it could be missing case reports and other grey literature contributions. Second, there is a risk of publication bias in a review of published studies. Third, due to the heterogeneity of the research methods across studies, we did not perform a meta-analysis to quantitatively examine how satellite imagery has been used to address the top NCDs in the world. This limits the depth of the analysis that could be achieved through more rigorous statistical exploration. Fourth, given that most studies are from developed regions, findings are skewed toward higher-income countries. Future studies should explore ways to include more diverse geographical inputs to research using satellite imagery in examining noncommunicable diseases.

Conclusions

Overall, this systematic review found satellite data to be an asset to NCD research. However, given the recent proliferation of satellites and the emerging capabilities of AI, using satellite imagery data to address the global health threat of NCDs has barely scratched the surface, particularly for locations most vulnerable to NCDs, such as low- and middle-income countries. Scientists and policymakers worldwide should take concerted and collaborative action to keep pace with the advancement of satellite imagery to produce better data-driven health outcomes.

Data availability

Data is provided within the manuscript or supplementary information files.

Abbreviations

AI:

Artificial intelligence

AOD:

Aerosol Optical Depth

BMI:

Body Mass Index

CNN:

Convolutional Neural Networks

DMSP:

Department of Meteorological Satellite Program

GWR:

Geographically weighted regression

KM:

Kilometer

LAN:

Light at night

M:

Meter

MAIAC:

Multi-angle Implementation of Atmospheric Correction

MODIS:

Moderate Resolution Imaging Spectroradiometer

NASA:

National Aeronautics and Space Administration

NCD:

Noncommunicable disease

NDVI:

Normalized Difference Vegetation Index

PM2.5:

Particulate matter 2.5

PRISMA:

Preferred Reporting Items for Systematic Reviews and Meta-analyses

t-SNE:

t-distributed Stochastic Neighbor Embedding

References

  1. World Health Organization. Noncommunicable Diseases. 2022. https://www.who.int/news-room/fact-sheets/detail/noncommunicable-diseases

  2. Waters H, Graf M. The Costs of Chronic Disease in the U.S.. Milken Institute; 2018 Aug [cited 2023 Sep 20]. https://milkeninstitute.org/sites/default/files/reports-pdf/ChronicDiseases-HighRes-FINAL_2.pdf

  3. Centers for Disease Control and Prevention. Centers for Disease Control and Prevention. 2023 [cited 2023 Aug 26]. PLACES: Local Data for Better Health. https://www.cdc.gov/places/index.html

  4. World health Organization. Environmental risk factors and NCDs. [cited 2024 Mar 21]. https://www.who.int/teams/noncommunicable-diseases/integrated-support/environmental-risk-factors-and-ncds

  5. World Health Organization W health statistics. The Global Health Observatory. 2023. WHO Indicator Data: Global Health Observatory. https://www.who.int/data/gho/data/themes/topics/indicator-groups/indicator-group-details/GHO/sdg-target-3.4-noncommunicable-diseases-and-mental-health#:~:text=Indicator Groups-,SDG Target 3.4 %7C Noncommunicable diseases and mental health%3A By 2030,mental health and well%2Dbeing

  6. Watkins DA, Msemburi WT, Pickersgill SJ, Kawakatsu Y, Gheorghe A, Dain K, et al. NCD countdown 2030: efficient pathways and strategic investments to accelerate progress towards the sustainable development goal target 3.4 in low-income and middle-income countries. Lancet. 2022;399(10331):1266–78.

    Article  Google Scholar 

  7. World Health Organization. Global NCD Compact 2020–2030. [cited 2023 Aug 9]. https://www.who.int/initiatives/global-noncommunicable-diseases-compact-2020-2030/achievements

  8. NASA Earth Science Data Systems. What is Remote Sensing? | Earthdata. 2019 [cited 2024 Mar 21]. https://www.earthdata.nasa.gov/learn/backgrounders/remote-sensing

  9. World Economic Forum. 2020 [cited 2023 Aug 29]. Who owns our orbit: Just how many satellites are there in space? https://www.weforum.org/agenda/2020/10/visualizing-easrth-satellites-sapce-spacex/

  10. Barnard PL, Vitousek S. Earth science looks to outer space. Nat Geosci. 2023;16(2):108–9.

    Article  CAS  Google Scholar 

  11. Yeh C, Perez A, Driscoll A, Azzari G, Tang Z, Lobell D, et al. Using publicly available satellite imagery and deep learning to understand economic well-being in Africa. Nat Commun. 2020;11(1):2583.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Amoroso N, Cilli R, Maggipinto T, Monaco A, Tangaro S, Bellotti R. Satellite data and machine learning reveal a significant correlation between NO2 and COVID-19 mortality. Environ Res. 2022;204:111970.

    Article  CAS  PubMed  Google Scholar 

  13. Desai AN, Kraemer MU, Bhatia S, Cori A, Nouvellet P, Herringer M, et al. Real-time epidemic forecasting: challenges and opportunities. Health Secur. 2019;17(4):268–75.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Rogers DJ, Randolph SE, Snow RW, Hay SI. Satellite imagery in the study and forecast of malaria. Nature. 2002;415(6872):710–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Jia P, Stein A, James P, Brownson RC, Wu T, Xiao Q, et al. Earth Observation: investigating Noncommunicable diseases from Space. Annu Rev Public Health. 2019;40(1):85–104.

    Article  PubMed  Google Scholar 

  16. Earthdata | Earthdata. [cited 2022 Nov 8]. https://www.earthdata.nasa.gov/

  17. ESA - Eduspace EN - Home. - What is remote sensing?. 2010 [cited 2024 Jan 25]. https://www.esa.int/SPECIALS/Eduspace_EN/SEMF9R3Z2OF_0.html#:~:text=Remote sensing is a way of collecting and,data being in direct contact with the object

  18. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg. 2021;88:105906.

    Article  PubMed  Google Scholar 

  19. Morgan RL, Whaley P, Thayer KA, Schünemann HJ. Identifying the PECO: a framework for formulating good questions to explore the association of environmental and other exposures with health outcomes. Environ Int. 2018;121(Pt 1):1027.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Howick J, Chalmers I, Glasziou P, Greenhalgh T, Heneghan C, Liberati A et al. OCEBM Levels of Evidence. Centre for Evidence-Based Medicine (CEBM), University of Oxford. 2021.

  21. Chan CC, Chuang KJ, Chen WJ, Chang WT, Lee CT, Peng CM. Increasing cardiopulmonary emergency visits by long-range transported Asian dust storms in Taiwan. Environ Res. 2008;106(3):393–400.

    Article  CAS  PubMed  Google Scholar 

  22. De Roos AJ, Kenyon CC, Yen YT, Moore K, Melly S, Hubbard RA, et al. Does living near Trees and other Vegetation affect the contemporaneous odds of Asthma Exacerbation among Pediatric Asthma patients? J Urban Health. 2022;99(3):533–48.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Evans J, Van Donkelaar A, Martin RV, Burnett R, Rainham DG, Birkett NJ, et al. Estimates of global mortality attributable to particulate air pollution using satellite imagery. Environ Res. 2013;120:33–42.

    Article  CAS  PubMed  Google Scholar 

  24. Gao J, Liu J, Xu R, Pandey S, Vankayala Siva VSKS, Yu D. Environmental Pollution Analysis and Impact Study—A Case Study for the Salton Sea in California. Atmosphere. 2022;13(6):914.

    Article  CAS  Google Scholar 

  25. Garzon-Chavez DR, Quentin E, Harrison SL, Parisi AV, Butler HJ, Downs NJ. The geospatial relationship of pterygium and senile cataract with ambient solar ultraviolet in tropical Ecuador. Photochem Photobiol Sci. 2018;17(8):1075–83.

    Article  CAS  PubMed  Google Scholar 

  26. Higgs G, Sterling DA, Aryal S, Vemulapalli A, Priftis KN, Sifakis NI. Aerosol Optical Depth as a measure of Particulate exposure using Imputed Censored Data, and relationship with Childhood Asthma Hospital Admissions for 2004 in Athens, Greece. Environ Health Insights. 2015;9s1:EHI.S15665.

  27. James P, Hart JE, Banay RF, Laden F. Exposure to greenness and mortality in a nationwide prospective cohort study of women. Environ Health Perspect. 2016;124(9):1344–52.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Nguyen HD, Azzi M, White S, Salter D, Trieu T, Morgan G, et al. The summer 2019–2020 wildfires in East Coast Australia and their impacts on Air Quality and Health in New South Wales, Australia. Int J Environ Res Public Health. 2021;18(7):3538.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Prabhu V, Shridhar V, Choudhary A. Investigation of the source, morphology, and trace elements associated with atmospheric PM10 and human health risks due to inhalation of carcinogenic elements at Dehradun, an Indo-Himalayan city. SN Appl Sci. 2019;1(5):429.

    Article  CAS  Google Scholar 

  30. Prud’homme G, Dobbin NA, Sun L, Burnett RT, Martin RV, Davidson A, et al. Comparison of remote sensing and fixed-site monitoring approaches for examining air pollution and health in a national study population. Atmos Environ. 2013;80:161–71.

    Article  Google Scholar 

  31. Qu Y, Yang B, Lin S, Bloom MS, Nie Z, Ou Y, et al. Associations of greenness with gestational diabetes mellitus: the Guangdong Registry of congenital heart Disease (GRCHD) study. Environ Pollut. 2020;266:115127.

    Article  CAS  PubMed  Google Scholar 

  32. Ramesh B, Jagger MA, Zaitchik BF, Kolivras KN, Swarup S, Yang B, et al. Estimating changes in emergency department visits associated with floods caused by Tropical Storm Imelda using satellite observations and syndromic surveillance. Health Place. 2022;74:102757.

    Article  PubMed  Google Scholar 

  33. Stowell JD, Geng G, Saikawa E, Chang HH, Fu J, Yang CE, et al. Associations of wildfire smoke PM2.5 exposure with cardiorespiratory events in Colorado 2011–2014. Environ Int. 2019;133:105151.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Zhang D, Bai K, Zhou Y, Shi R, Ren H. Estimating ground-level concentrations of Multiple Air Pollutants and their Health impacts in the Huaihe River Basin in China. Int J Environ Res Public Health. 2019;16(4):579.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Allen RW, Gombojav E, Barkhasragchaa B, Byambaa T, Lkhasuren O, Amram O, et al. An assessment of air pollution and its attributable mortality in Ulaanbaatar, Mongolia. Air Qual Atmos Health. 2013;6(1):137–50.

    Article  CAS  PubMed  Google Scholar 

  36. Dagliati A, Marinoni A, Cerra C, Decata P, Chiovato L, Gamba P, et al. Integration of administrative, clinical, and Environmental Data to support the management of type 2 diabetes Mellitus: from satellites to Clinical Care. J Diabetes Sci Technol. 2016;10(1):19–26.

    Article  Google Scholar 

  37. Eldeirawi K, Kunzweiler C, Zenk S, Finn P, Nyenhuis S, Rosenberg N, et al. Associations of urban greenness with asthma and respiratory symptoms in Mexican American children. Ann Allergy Asthma Immunol. 2019;122(3):289–95.

    Article  PubMed  Google Scholar 

  38. Fan J, Guo Y, Cao Z, Cong S, Wang N, Lin H, et al. Neighborhood greenness associated with chronic obstructive pulmonary disease: a nationwide cross-sectional study in China. Environ Int. 2020;144:106042.

    Article  PubMed  Google Scholar 

  39. Jimenez MP, Oken E, Gold DR, Luttmann-Gibson H, Requia WJ, Rifas-Shiman SL, et al. Early life exposure to green space and insulin resistance: an assessment from infancy to early adolescence. Environ Int. 2020;142:105849.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Klompmaker JO, Laden F, Browning MHEM, Dominici F, Ogletree SS, Rigolon A, et al. Associations of parks, greenness, and blue space with cardiovascular and respiratory disease hospitalization in the US Medicare cohort. Environ Pollut. 2022;312:120046.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Lambert KA, Lodge C, Lowe AJ, Prendergast LA, Thomas PS, Bennett CM, et al. Pollen exposure at birth and adolescent lung function, and modification by residential greenness. Allergy. 2019;74(10):1977–84.

    Article  PubMed  Google Scholar 

  42. Qazi S, Iqbal J, Khan JA. Assessment of the health impact of paper mulberry (Broussonetia papyrifera L.), an invasive plant species in Islamabad, Pakistan. Geospatial Health. 2019 Nov 12 [cited 2023 Jul 12];14(2). https://geospatialhealth.net/index.php/gh/article/view/727

  43. Silveira IHD, Junger WL. Green spaces and mortality due to cardiovascular diseases in the city of Rio De Janeiro. Rev Saúde Pública. 2018;52:49.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Vargas-Cuentas NI, Roman-Gonzalez A, Mantari AA, Muñoz LA. Chagas disease study using satellite image processing: a Bolivian case. Acta Astronaut. 2018;144:216–24.

    Article  Google Scholar 

  45. Walker BB, Brinkmann ST, Große T, Kremer D, Schuurman N, Hystad P, et al. Neighborhood Greenspace and Socioeconomic Risk are Associated with Diabetes Risk at the Sub-neighborhood Scale: results from the prospective Urban and Rural Epidemiology (PURE) Study. J Urban Health. 2022;99(3):506–18.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Bauer SE, Wagner SE, Burch J, Bayakly R, Vena JE. A case-referent study: light at night and breast cancer risk in Georgia. Int J Health Geogr. 2013;12(1):23.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Kloog I, Haim A, Stevens RG, Barchana M, Portnov BA. Light at night co-distributes with incident breast but not Lung Cancer in the Female Population of Israel. Chronobiol Int. 2008;25(1):65–81.

    Article  PubMed  Google Scholar 

  48. Medgyesi DN, Trabert B, Fisher JA, Xiao Q, James P, White AJ, et al. Outdoor light at night and risk of endometrial cancer in the NIH-AARP diet and health study. Cancer Causes Control. 2023;34(2):181–7.

    Article  PubMed  Google Scholar 

  49. Park Y, Ramirez Y, Xiao Q, Liao LM, Jones GS, McGlynn KA. Outdoor light at night and risk of liver cancer in the NIH-AARP diet and health study. Cancer Causes Control. 2022;33(9):1215–8.

    Article  PubMed  Google Scholar 

  50. Portnov BA, Stevens RG, Samociuk H, Wakefield D, Gregorio DI. Light at night and breast cancer incidence in Connecticut: an ecological study of age group effects. Sci Total Environ. 2016;572:1020–4.

    Article  CAS  PubMed  Google Scholar 

  51. Xiao Q, James P, Breheny P, Jia P, Park Y, Zhang D, et al. Outdoor light at night and postmenopausal breast cancer risk in the NIH-AARP diet and health study. Int J Cancer. 2020;147(9):2363–72.

    Article  CAS  PubMed  Google Scholar 

  52. Xiao Q, Jones RR, James P, Stolzenberg-Solomon RZ. Light at night and risk of pancreatic Cancer in the NIH-AARP Diet and Health Study. Cancer Res. 2021;81(6):1616–22.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Hidalgo-García D, Arco-Díaz J. Spatiotemporal analysis of the surface urban heat island (SUHI), air pollution and disease pattern: an applied study on the city of Granada (Spain). Environ Sci Pollut Res. 2023;30(20):57617–37.

    Article  Google Scholar 

  54. Yuan Q, Shen H, Li T, Li Z, Li S, Jiang Y, et al. Deep learning in environmental remote sensing: achievements and challenges. Remote Sens Environ. 2020;241:111716.

    Article  Google Scholar 

  55. Upegui E, Viel JF. GeoEye Imagery and Lidar Technology for small-area Population Estimation: an epidemiological viewpoint. Photogramm Eng Remote Sens. 2012;78(7):693–702.

    Article  Google Scholar 

  56. Liu Y, Zhao B, Cheng Y, Zhao T, Zhang A, Cheng S, et al. Does the quality of street greenspace matter? Examining the associations between multiple greenspace exposures and chronic health conditions of urban residents in a rapidly urbanising Chinese city. Environ Res. 2023;222:115344.

    Article  CAS  PubMed  Google Scholar 

  57. Bibault JE, Bassenne M, Ren H, Xing L. Deep learning prediction of Cancer Prevalence from Satellite Imagery. Cancers. 2020;12(12):3844.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Maharana A, Nsoesie EO. Use of Deep Learning to Examine the Association of the built Environment with Prevalence of Neighborhood adult obesity. JAMA Netw Open. 2018;1(4):e181535.

    Article  PubMed  PubMed Central  Google Scholar 

  59. Chatfield K, Simonyan K, Vedaldi A, Zisserman A. Return of the Devil in the Details: Delving Deep into Convolutional Nets. arXiv; 2014 [cited 2023 Sep 14]. http://arxiv.org/abs/1405.3531

  60. Kim Y, Bak SH, Kwon SO, Kim H, Kim WJ, Lee CY. Association between Long-Term exposure to PM2.5 and lung imaging phenotype in CODA Cohort. Atmosphere. 2021;12(2):282.

    Article  Google Scholar 

  61. Brown SC, Lombard J, Wang K, Byrne MM, Toro M, Plater-Zyberk E, et al. Neighborhood Greenness and Chronic Health conditions in Medicare Beneficiaries. Am J Prev Med. 2016;51(1):78–89.

    Article  PubMed  Google Scholar 

  62. Gariepy G, Kaufman JS, Blair A, Kestens Y, Schmitz N. Place and health in diabetes: the neighbourhood environment and risk of depression in adults with type 2 diabetes. Diabet Med. 2015;32(7):944–50.

    Article  CAS  PubMed  Google Scholar 

  63. Wang K, Lombard J, Rundek T, Dong C, Gutierrez CM, Byrne MM, et al. Relationship of Neighborhood Greenness to Heart Disease in 249 405 US Medicare Beneficiaries. J Am Heart Assoc. 2019;8(6):e010258.

    Article  PubMed  PubMed Central  Google Scholar 

  64. Yitshak Sade M, Novack V, Ifergane G, Horev A, Kloog I. Air Pollution and ischemic stroke among young adults. Stroke. 2015;46(12):3348–53.

    Article  CAS  PubMed  Google Scholar 

  65. Yuan Y, Wu Y, Zhao H, Ren J, Su W, Kou Y, et al. Tropospheric formaldehyde levels infer ambient formaldehyde-induced brain diseases and global burden in China, 2013–2019. Sci Total Environ. 2023;883:163553.

    Article  CAS  PubMed  Google Scholar 

  66. Bureau UC. Census.gov. [cited 2023 Aug 26]. Census.gov. https://www.census.gov/en.html

  67. Schwalbe N, Wahl B. Artificial intelligence and the future of global health. Lancet. 2020;395(10236):1579–86.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  68. Lehnert P, Niederberger M, Backes-Gellner U, Bettinger E. T Jaworski editor 2023 Proxying economic activity with daytime satellite imagery: filling data gaps across time and space. PNAS Nexus 2 4 pgad099.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Rolf E, Proctor J, Carleton T, Bolliger I, Shankar V, Ishihara M, et al. A generalizable and accessible approach to machine learning with global satellite imagery. Nat Commun. 2021;12(1):4392.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  70. Malekzadeh A, Michels K, Wolfman C, Anand N, Sturke R. Strengthening research capacity in LMICs to address the global NCD burden. Glob Health Action. 2020;13(1):1846904.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Canadian Space Agency, Canadian Space A. 2018 [cited 2023 Aug 11]. How satellites help you stay healthy. https://www.asc-csa.gc.ca/eng/satellites/everyday-lives/how-satellites-help-you-stay-healthy.asp

  72. mosaiks.org. 2023 [cited 2023 Aug 11]. mosaiks.org. https://www.mosaiks.org

  73. National Oceanic and Atmospheric Administration. GOES-R Algorithm Theoretical Basis Documents. 2023 [cited 2023 Dec 3]. https://www.star.nesdis.noaa.gov/goesr/documentation_ATBDs.php

  74. NASA. Worldview. [cited 2023 Sep 17]. https://worldview.earthdata.nasa.gov/

  75. Earth Online. [cited 2023 Sep 17]. https://earth.esa.int/eogateway/

  76. EUMETSAT. EUMETSAT | Monitoring the weather and climate from space | EUMETSAT. [cited 2023 Sep 17]. https://www.eumetsat.int/

  77. Gettelman A, Geer AJ, Forbes RM, Carmichael GR, Feingold G, Posselt DJ, et al. The future of Earth system prediction: advances in model-data fusion. Sci Adv. 2022;8(14):eabn3488.

    Article  PubMed  PubMed Central  Google Scholar 

  78. Chen CP, Zhang CY. Data-intensive applications, challenges, techniques and technologies: a survey on Big Data. Inf Sci. 2014;275:314–47.

    Article  Google Scholar 

  79. Kotawadekar R. 9 - Satellite data: big data extraction and analysis. In: Binu D, Rajakumar BR, editors. Artificial Intelligence in Data Mining. Academic Press; 2021 [cited 2023 Aug 26]. pp. 177–97. https://www.sciencedirect.com/science/article/pii/B9780128206010000082

  80. Abu Qdais H, Shatnawi N. Assessing and predicting landfill surface temperature using remote sensing and an artificial neural network. Int J Remote Sens. 2019;40(24):9556–71.

    Article  Google Scholar 

  81. Baghanam AH, Vakili AT, Nourani V, Dąbrowska D, Soltysiak M. AI-based ensemble modeling of landfill leakage employing a lysimeter, climatic data and transfer learning. J Hydrol. 2022;612:128243.

    Article  CAS  Google Scholar 

  82. Tahir A, Munawar HS, Akram J, Adil M, Ali S, Kouzani AZ, et al. Automatic target detection from Satellite Imagery using machine learning. Sensors. 2022;22(3):1147.

    Article  PubMed  PubMed Central  Google Scholar 

  83. Tariq A, Siddiqui S, Sharifi A, Shah SHIA. Impact of spatio-temporal land surface temperature on cropping pattern and land use and land cover changes using satellite imagery, Hafizabad District, Punjab, Province of Pakistan. Arab J Geosci. 2022;15(11):1045.

    Article  Google Scholar 

  84. Jiang H, Peng M, Zhong Y, Xie H, Hao Z, Lin J, et al. A Survey on Deep Learning-based change detection from high-resolution remote sensing images. Remote Sens. 2022;14(7):1552.

    Article  Google Scholar 

  85. Bistron M, Piotrowski Z. Artificial Intelligence Applications in Military systems and their influence on sense of security of citizens. Electronics. 2021;10(7):871.

    Article  Google Scholar 

  86. United Nations Register of Objects. Launched into Outer Space. [cited 2023 Aug 29]. https://www.unoosa.org/oosa/en/spaceobjectregister/index.html

Download references

Acknowledgements

The authors would like to thank Lei Hua for his review of the manuscript.

Funding

The authors received no external financial support for the research, authorship, and/or publication of this article.

Author information

Authors and Affiliations

  1. Shirley Ryan AbilityLab, 355 East Erie Street, Chicago, IL, 60611, USA

    Elizabeth J. Folkmann

  2. School of Health Studies, College of Health and Human Sciences, Northern Illinois University, 209 Wirtz Hall, DeKalb, IL, 60115, USA

    M. Courtney Hughes

  3. College of Business, Northern Illinois University, 328 Barsema Hall, DeKalb, IL, USA

    Uzma Amzad Khan

  4. College of Engineering and Engineering Technology, Northern Illinois University, 590 Garden Road, DeKalb, IL, USA

    Mahdi Vaezi

Authors
  1. Elizabeth J. Folkmann
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. M. Courtney Hughes
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Uzma Amzad Khan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Mahdi Vaezi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Review conception and design: MV, MCH, EF, UK. Search strategy and literature search: EF, UK. Study coding, selection, and data extraction: EF, UK. Review and interpretation: EF, MCH. Drafting of manuscript: EF, MCH. Critical revision: EF, MCH, MV. All authors read and approved the final manuscript.

Corresponding author

Correspondence to M. Courtney Hughes.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Folkmann, E.J., Hughes, M.C., Khan, U.A. et al. Examining noncommunicable diseases using satellite imagery: a systematic literature review. BMC Public Health 24, 2774 (2024). https://doi.org/10.1186/s12889-024-20316-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12889-024-20316-z

Keywords

BMC Public Health

ISSN: 1471-2458

Contact us