Cardiovascular disease stands as the foremost cause of mortality across the globe. Cardiac magnetic resonance imaging (CMR), widely hailed as the 'gold standard' for diagnosing such conditions, has yet to fully unlock its clinical promise. This is primarily due to the hurdles posed by the need for a high degree of expertise in interpreting images and the unequal distribution of high-quality medical resources. To tackle this issue, the Beijing Academy of Artificial Intelligence (BAAI), in partnership with hospitals, has introduced the industry's pioneering multimodal agent for cardiac MRI—the BAAI Cardiac Agent.

Leveraging an Agent-Expert framework, this agent facilitates closed-loop management throughout the entire diagnostic and therapeutic journey. It automates the end-to-end analysis process and generates standardized reports. Extensive evaluations conducted by the research team reveal a remarkable alignment between its diagnostic outcomes and the judgments of clinical experts.

The creation and open-sourcing of the BAAI Cardiac Agent signify a pivotal advancement for AI in the realm of complex medical imaging diagnosis and treatment. It aids in enhancing interpretation efficiency and minimizing diagnostic discrepancies. Presently, the core code and evaluation datasets of this agent are entirely accessible to the public, fostering collaborative innovation and the broad adoption of AI technology within the medical sector.