Presented as a Late-breaking science at Heart Failure 2025 on May 18 in Belgrade, Serbia, the Heart2Miss screening programme conducted in Malaysia drew significant attention for its innovative approach to early heart failure detection. Images were acquired by bioscience graduates trained through a focused four-week program in handheld ultrasound and AI-assisted analysis (Us2.ai).
Using a decentralized, community-based ultrasound triage model, the programme successfully screened 1,000 high-risk diabetic patients — demonstrating that heart failure can be identified earlier, even within primary care settings. By equipping bioscience graduates to serve as mobile sonographers, the initiative reduced pressure on tertiary centers, streamlined patient referrals, and opened new career pathways in healthcare diagnostics.
Key Findings
- The Heart2Miss programme successfully screened 1,000 high-risk diabetic patients for early heart failure using a decentralised, community-based AI echo triage model, demonstrating that large-scale cardiac screening is feasible outside tertiary centres.
- Bioscience graduates completed a focused 4-week training programme in handheld ultrasound and AI-assisted analysis, and were able to acquire diagnostic-quality echocardiographic images, showing that task-shifting to non-specialist operators is achievable with structured training.
- The programme reduced pressure on tertiary imaging centres by enabling community-level triage and streamlining referral pathways, with most patients managed without needing to travel to a specialist centre.
- Heart2Miss offers a scalable, low-cost model for early heart failure detection in resource-constrained settings, with potential applicability across other regions facing similar challenges in cardiovascular diagnostic access.
The success of Heart2Miss highlights the powerful potential of combining AI with community-based care — representing a significant leap forward in early diagnosis and equitable access to cardiovascular care in Malaysia, and offering a scalable model for other resource-constrained clinical settings.
*Heart2Miss was an initiative established with public-private partnerships in local community diabetes care. This study was made possible with the support of study grants from AstraZeneca and technical support and automated echocardiography software provided by Us2.ai.*