This project engages with the proliferation of abusive and biased stereotypes from colonial and humanitarian photography through generative AI technology, and investigates their consequences for science and society. This is a pressing issue: AI simultaneously absorbs and learns from real images, which, in the case of global health, have been marked by racism, coloniality and sexism, meaning that given images become a cluster for generative AI to learn from biased depictions and perpetuate negative stereotypes. Such cycles have to be studied and eliminated in order to move toward more equal postcolonial societies and promote a culture of value-sensitive depictions of vulnerable people. The project builds and greatly expands on the emerging methodology of purposeful generation and value-sensitive evaluation of AI-generated Global Health visuals, recently pioneered by Prof. Koen Peeters (the supervisor) and Dr. Alenichev, and encapsulated in a Lancet Global Health Article in August 2023. Offering a first-ever systematic study of AI-generated Global Heath visuals, AIrbrush sets five core objectives and asks: How should the international community account for generative AI as part of the internationally set goal of decolonizing Global Health and its visual culture, and tackle biased depictions of race, class, gender, and other socially enacted markers of similarity and difference? AIrbrush answers this question by analysing the substrate of the real global health images AI learns from, evaluates the learning progress and the reproduction and modification of such tropes by AI, theorizes this relationship and outlines societal outcomes with regard to the future of respectful depictions in the AI era. The findings from this study will be encapsulated as academic articles, a thematic webinar, a collaboration with the WHO AI and Ethics research group, and an art exhibition at ITM (the host) and in other places, among other outputs.

With chest X-ray and molecular tests virtually absent at the primary healthcare care level, where most patients with presumptive TB present in sub-Saharan Africa (SSA), there is a need for accessible, affordable and scalable diagnostic tools for TB triage. CAD LUS4TB represents an interdisciplinary partnership spanning across Western (Francophone) and Southern-African regions with EU countries, aimed at enhancing access to effective TB triage to rule out TB disease among symptomatic adult patients presenting at the primary healthcare level. This initiative focuses on generating population-tailored evidence and advocating for the integration of computer-assisted diagnosis (CAD) using artificial intelligence (AI) to support the implementation of lung ultrasound (LUS) into healthcare policy. Unlike typical vertical triage tests, US has multiple other existing AI-assisted diagnostic tools and can facilitate a multi disease approach after TB exclusion, including for pneumonia and cardiovascular assessment. We propose to externally validate and deploy a novel digital technology adapting image-based analysis tools and software for mobile phone ultrasound applications. AI technology sharing serves as one of its key pillars. The adoption of CAD-LUS requires a comprehensive, interdisciplinary, translational approach to clinical research. Our consortium comprises these key fields, including clinical research, diagnostics, implementation science, social science, health economics and policy translation, as well as data/computer science. It addresses all expected outcomes and contributes to several specific expected impacts of this call. Evidence on the integration of CAD-LUS is expected to accelerate adoption of accessible triage tools for TB in SSA and support achieving target 3.3 of the Sustainable Development Goals. The CAD LUS4TB tool is anticipated to achieve a high diagnostic yield due to its user-friendliness, scalability and possibility to address multiple diseases.

Gambiense human African trypanosomiasis (gHAT) is a neglected tropical disease caused by trypanosome parasites. gHAT is fatal if left untreated. So far, treatment options for gHAT were limited and toxic, forcing control programs to avoid overtreatment through complex diagnostic procedures, including screening with a serological test, laborious microscopic confirmation of seropositives and lumbar puncture for disease stage determination. This resulted in loss of up to 50% of gHAT cases, which remained untreated. Recently a non-toxic single dose oral drug, acoziborole, has shown 98.1% efficacy in a phase III trial, irrespective of gHAT disease stage. Acoziborole removes the need for lumbar puncture and appears safe enough to treat serological suspects without microscopic confirmation (Screen & treat). The STROGHAT project 1° will evaluate effectiveness of a Screen & treat approach to rapidly reduce gHAT prevalence in an entire focus; 2° will extend acoziborole safety documentation; 3° and will analyze costs of this new approach. To achieve these objectives, Screen & treat will be implemented, actively and passively, for 3 consecutive years in the gHAT focus of Nord Equateur in D.R. Congo. Available geographical information will be exploited to specifically target villages where gHAT was recently, or still is present. Detection at a reference laboratory, of the trypanosomes nucleic acids in blood collected before treatment, will retrospectively identify true gHAT cases among the treated serological suspects. After 3 years of intervention, the gHAT prevalence in the focus will be re-estimated. STROGHAT intends to provide the first evidence for recommending Screen & treat to national HAT control programs for elimination of gHAT. Through facilitated diagnosis, increased acceptability and access to treatment, STROGHAT will contribute to achieving the goal of stopping gHAT transmission by 2030, as defined by the World Health Organization.