Top 16%
dans ce catalogue (246 pratiques notées)
Les scores se concentrent dans le haut de l’échelle ; la position au sein du catalogue est donc souvent plus parlante que le chiffre seul.
Evidence of impact / measured public value3/3
A peer-reviewed study reports a quantified AUC-ROC of 0.86 and 97.7% sensitivity against a bacteriological reference across a real 8,386-person screening cohort, the strongest evidence base of this batch.
Transparency, fairness & accountability2/3
Accuracy and sensitivity metrics were published in a peer-reviewed journal and corroborated by WHO reporting, though no source details the CAD algorithm's training data or internal governance.
Transferability / demonstrated replication2/3
The WHO/Global Fund-backed CAD approach used in Cambodia is part of a model deployed across multiple countries' TB programmes, though this specific study evidences only the Cambodia deployment.
Scalability beyond pilot2/3
CENAT's portable AI-CAD units are used routinely in provincial active case-finding between national surveys, beyond a single pilot, though no nationwide throughput total is documented.
Governance, capability & sustainability2/3
CENAT operates the programme under the Ministry of Health with sustained WHO, Global Fund, USAID and JATA partnership spanning over two decades of TB programming.
Evidoria. Détection assistée par IA pour le dépistage de la tuberculose — Le programme de radiographie thoracique du CENAT au Cambodge. ID persistant: a3696305-7f42-4b5d-8432-a27ec5306f66.