Eko Health, in collaboration with the G-ACT Foundation, has announced transformative findings from the DAMSUN-HF (Detection and Management of Heart Failure with SENSORA in Underserved Nations – Heart Failure) study. The prospective validation demonstrated that AI-assisted digital auscultation can accurately identify heart failure with reduced ejection fraction (HFrEF) at the point of care in Sub-Saharan Africa, marking a major advance in equitable cardiac detection.
The results, published in Circulation and presented as a Late-Breaking Clinical Trial at the American Heart Association Scientific Sessions, show that Eko’s SENSORA® platform achieved 97% sensitivity and 94% negative predictive value in identifying patients with LVEF ≤ 40%, with 76% specificity. Performance remained strong across age and sex groups, supporting consistent frontline use.
DAMSUN-HF enrolled 115 adults with cardiopulmonary symptoms across Ghana’s district “spoke” clinics and tertiary “hub” centers. AI-assisted auscultation was compared with blinded transthoracic echocardiography (TTE). The study also evaluated real-world implementation, with more than 95% of participants completing every protocol step from early community screening to specialist review. Over 90% of AI-flagged cases were assessed by a cardiologist within 48 hours, demonstrating seamless integration into Ghana’s hub-and-spoke healthcare system without the need for advanced imaging.
The collaboration highlights the potential for AI-enabled auscultation to bridge diagnostic gaps in underserved regions by enabling rapid triage and earlier intervention in settings where echocardiography and cardiology resources are limited.
The study was supported by the G-ACT Foundation and conducted with Eko Health under the oversight of the Komfo Anokye Teaching Hospital Institutional Review Board. Findings reinforce a scalable model connecting community clinics to tertiary centers through cloud-based data sharing and expedited specialist evaluation, advancing access to early cardiac disease detection.
Also Read
