Page 88 - SAMRC AnnualReport 2025-26

proteomic and clinical data. These methods are used the interpretation of digital chest radiographs for
to identify biomarker signatures associated with tuberculosis screening in high-burden communities.
disease severity, infectiousness, treatment response, Building on this work, new studies are evaluating
and antimicrobial resistance across blood, sputum, AI-enabled digital stethoscopes that analyse
and urine, enabling multimodal biomarker discovery lung sounds together with portable digital lung
and a more comprehensive, data-driven approach to ultrasound to determine their utility for TB diagnosis.
diagnosis and risk stratification.
These technologies have the potential to enhance
AI is also being applied to community-based case screening accuracy, expand access to diagnostic
finding and screening. The XACT-19 study uses services in resource-limited settings and enable
computer-aided detection (CAD) software to support earlier referral for confirmatory testing and treatment.

86 SA M R C A N N U A L R EP O R T 2 0 2 5 / 2 0 2 6