Workflow
肝包虫AI辅助诊断产品
icon
Search documents
开立医疗AI首席科学家周国义:医疗智能的最高水平是人机共检
Core Viewpoint - The highest level of medical intelligence is considered to be L4, which emphasizes "human-machine collaboration" rather than a fully automated medical environment [1][20]. Industry Insights - The global medical AI market is projected to exceed $100 billion by 2030, with medical imaging AI being one of the fastest-growing segments [2]. - The company has been a pioneer in systematically laying out ultrasound AI since 2016, adopting a "device + AI" strategy that integrates AI deeply into medical devices [3][4]. Technological Development - The company focuses on three core advantages of its "device + AI" approach: implementation advantage, integration advantage, and performance advantage [4]. - The "Fengyan S-Fetus" technology, which integrates ultrasound technology with AI, has achieved significant milestones, including obtaining the first domestic prenatal ultrasound AI medical device certification [4][11]. Data and Model Training - High-quality, well-distributed data is crucial for the effective application of AI in clinical settings [5][15][16]. - The company collaborates with various medical institutions to build a clinical data cooperation system, ensuring compliance with ethical standards and data privacy [16]. Future Directions - The company aims to achieve a "fully intelligent machine" concept, where AI drives the entire clinical process, covering major application areas such as gynecology, obstetrics, cardiology, and more [21]. - The integration of visual models with large language models is being explored to enhance diagnostic capabilities [17]. Regulatory and Policy Considerations - The registration process for AI products is more stringent than for traditional medical devices, emphasizing the need for rigorous clinical trials and data integrity [12][11]. - There is a call for improved regulatory frameworks to support the integration of AI in healthcare, ensuring that AI can be effectively utilized while maintaining human oversight [22][23].