Détecter la tuberculose par algorithmes : les valorisations d’une technologie de surveillance par la santé mondiale

Abstract

Computer-aided detection software based on artificial intelligence (CAD-IA), combined with digital chest X-rays, has recently been touted as an easy solution to a complex problem: “ending tuberculosis by 2030”. The WHO recommended the use of these devices in 2021, and numerous public/private partnerships have helped to evaluate them, creating a market for these first artificial intelligence-based global health tools. This article explores the stages in the creation of value and a market for these detection algorithms, thanks to an “accelerationist” regime that is well entrenched within the field of global health. By following the social life of the data used to develop, validate, sell, and profit from CAD-IA devices, we analyze how a new form of technological TB detection used in global health, despite its disruptive claims, remains caught up in processes of inequality that drive TB epidemics in many contexts and are indirectly valued scientifically, economically, and politically, by the technology itself.

Keywords

• surveillance
• global health
• capitalism
• Tuberculosis
• algorithms
• artificial intelligence