Singapore Leverages AI to Enhance Diagnostics in Resource-Limited Healthcare

image credit- freepik

Following a cardiac arrest, families and healthcare professionals often grapple with excruciating uncertainty regarding a patient’s recovery prospects. This dilemma is amplified in hospitals with limited resources, where access to advanced diagnostic tools and extensive datasets is severely restricted.

In a significant step forward, researchers from Duke-NUS Medical School in Singapore, along with their partners, have modified an advanced AI model to accurately forecast neurological recovery post-cardiac arrest within resource-limited settings.

This groundbreaking research, published in npj Digital Medicine, utilized transfer learning—an innovative AI technique that adapts pre-trained models from large datasets to new environments with minimal local data. This approach enhances the model’s efficacy without necessitating extensive data collection, making it especially applicable to low- and middle-income countries.

While AI has the potential to transform healthcare delivery, establishing suitable governance frameworks is crucial for its safe and ethical use. Current regulations for medical technologies often overlook AI-specific challenges, such as privacy risks and model inaccuracies, and do not sufficiently clarify accountability for the safe implementation and monitoring of these tools.

To bridge these regulatory gaps, the research team at Duke-NUS has proposed an international consortium known as the Partnership for Oversight, Leadership, and Accountability in Regulating Intelligent Systems-Generative Models in Medicine (POLARIS-GM).

The consortium’s mission is to develop practical guidance for overseeing new tools, assessing their impact, setting safety standards, and adapting them for use in resource-constrained environments.