Project Title: Using artificial intelligence to detect high-casualty epidemics from the satellite imagery of burial sites
Project Description: Detecting high-casualty epidemics is essential for health authority to prospectively prevent epidemics and to retrospectively address health needs. Prospectively identifying epidemics can enable timely containment and mitigation effort, saving lives and potentially averting larger-scale epidemics and pandemics. Retrospectively, identifying locations that have been hard-hit by epidemics can help mobilize resources to address their long-term consequences. Epidemics can increase cancer, diabetes, neurodevelopmental disorders, and long-term sequelae, as well as trauma, bereavements, orphan hood, poverty, and food... Detecting high-casualty epidemics is essential for health authority to prospectively prevent epidemics and to retrospectively address health needs. Prospectively identifying epidemics can enable timely containment and mitigation effort, saving lives and potentially averting larger-scale epidemics and pandemics. Retrospectively, identifying locations that have been hard-hit by epidemics can help mobilize resources to address their long-term consequences. Epidemics can increase cancer, diabetes, neurodevelopmental disorders, and long-term sequelae, as well as trauma, bereavements, orphan hood, poverty, and food insecurity, all of which require planning and resources. Unfortunately, many low- and middle-income countries (LMICs) lack data systems to detect high casualty epidemics, either prospectively or retrospectively. LMICs lack reliable and timely reporting of healthcare utilization and morality in healthcare settings, a challenge floodlit by vast under-reporting of patients presenting and dying with COVID-19. In some cases, no COVID-19 data were reported at all: for example, Tanzania has not officially reported COVID-19 statistics since may 2020. Satellite imagery is emerging as an important data source for LMICs because it does not rely on local data collection and reporting infrastructure. Satellite imagery of burial sites has been used to detect high-casualty epidemics. however, application of this method have been limited in time and space because manually labeling graves in satellite imagery are too laborious to conduct at scale. advances in artificial intelligence, particularly convolutional neural network for machine vision, could automate burial site image processing for continuous global-scale detection
Principal Investigator : Samson Kiware
Department Name :
Time frame: (2022-09-12) - (2025-08-31)
Nationa Institutes of Health (Prime)
Nationa Institutes of Health (Prime)
Nationa Institutes of Health (Prime)
External Collaborating Partners
None added yet.