The study, thought to be the first to use machine-learning to measure housingconditions in sub-Saharan Africa, has found housing qualitytransformed but the persistence of slum conditions compromise health.
Medical entomologist, science prodigy, and Associate Professor of Public Health at Wits University, Fredros Okumu, was 18 years old when he was recruited as live mosquito bait as part of a Ph.D. research experiment. Inside that tent waiting for the insanity-invoking buzz near his ears, Okumu’s fate was sealed: he’d work tirelessly during his academic career to find a solution to one of humankind’s most vexing public health problems in Africa—the deaths caused by malaria.
His initial focus was inside the home: he simulated the attractive qualities of human beings to mosquitoes (such as blood, scent and other particular biological markers) in a “decoy site,” which contained pathogenic fungi to kill the flying critters. This complemented other interventions like nets and insecticides. In 2009, his team developed location models to determine where best to place mosquito decoy devices using digital geographical information systems and participatory community mapping.