【2025医疗人工智能报告】：价值计量与支付探索，医疗人工智能的两个困局

Core Insights - The medical AI industry is experiencing high growth despite not yet achieving scalable profitability, with the Chinese solutions market projected to grow from 16.4 billion yuan in 2024 to 35.3 billion yuan by 2030, reflecting a CAGR of 13.63% [1] - Significant changes in medical AI by 2025 include breakthroughs in large models and increased participation from medical institutions [1] - The deployment of large models in hospitals is accelerating, with all top 100 hospitals in China having completed large model deployments by May 2025, and 38 hospitals developing 55 vertical medical models tailored to their needs [1] Market Growth - The medical AI market in China is expected to expand significantly, with a projected market size of 35.3 billion yuan by 2030 [1] - The integration of various disciplines such as computer science, industrial engineering, and medicine is driving the growth of medical AI [1] Technological Advancements - The introduction of DeepSeek-R1 has lowered the entry barriers for large models, prompting hospital administrators to actively deploy necessary infrastructure [1] - Innovations such as parameter-efficient fine-tuning (PEFT) and mixture of experts (MoE) are enhancing the capabilities of large models [1] Doctor Engagement - Doctors are showing greater enthusiasm for practical applications of large models compared to traditional AI, with some circumventing procurement restrictions to continue research [2] - Over 90% of doctors who have used related AI tools report positive feedback, indicating that AI can enhance surgical precision and reduce complication rates [4] Policy Support - Recent policies are increasingly supportive of AI in healthcare, aiming to establish high-quality data sets and trusted data spaces by 2027 [6] - The implementation of guidelines for AI and medical applications is expected to create a conducive environment for the development of large models [6] Challenges in Commercialization - The value generated by AI in different deployment environments is inconsistent, making it difficult for hospitals to accurately assess benefits and hindering commercialization [7] - Short-term interests of hospitals and doctors often conflict, with AI deployment benefiting doctors but not necessarily translating to immediate hospital gains [8] Long-term Perspectives - In the long term, improved surgical quality through AI could enhance hospital reputation and attract more patients, benefiting both departments and doctors [10] - AI's ability to save time for doctors may lead to increased research opportunities, enhancing both individual and institutional capabilities [11] Specialty Focus: Thoracic Surgery - Thoracic surgery has a high demand for AI to improve operational efficiency and reduce redundant diagnostics [16] - AI applications in thoracic surgery have shown significant efficiency improvements, with diagnostic times reduced by up to 84% in some cases [18] - The introduction of AI in complex surgical planning has been shown to optimize procedures and reduce risks associated with needle placement [19] Data Governance and Assetization - The establishment of data as a production factor is accelerating the exploration of data assetization in healthcare, with a focus on efficient data governance and reuse [27] - The development of trusted data spaces is crucial for facilitating secure data sharing among healthcare stakeholders, promoting deeper integration and utilization of medical data [30]