After a cardiac arrest, uncertainty about recovery rates can overwhelm families and healthcare providers, especially in resource-limited settings. Researchers from Duke-NUS Medical School have developed an advanced AI model to predict neurological recovery in these environments, utilizing transfer learning to adapt pre-trained models without needing extensive local data. Their research, published in npj Digital Medicine, demonstrates improved accuracy in Vietnam, predicting outcomes in 80% of cases compared to 46% with the original model.
Additionally, AI applications in low- and middle-income countries show promise, enhancing diagnostics and decision-making, such as chatbots providing information in South Africa and smartphone apps detecting malaria in Sierra Leone. However, significant barriers persist, including limited infrastructure and expertise. Researchers propose the POLARIS-GM consortium to create a governance framework to ensure safe and ethical AI use in healthcare, promoting best practices and global consensus on AI regulation.
For more information, refer to: Li et al., npj Digital Medicine, 2025.
Source link
