转自：连线Insight 文 | 连线Insight 王慧莹 编辑 | 子夜 编者按： 岁末将至，站在2025年的时间节点回望，技术浪潮的奔涌、消费需求的变迁、商业模式的迭代，构成了全新的商业图景。连线Insight推出年终盘点专题系 列，试图捕捉不同企业在这幅变局图景中，如何应对挑战、抓住机遇。本期为第一篇，关注大健康行业。 当人们谈论"健康"，人们究竟在谈论什么？ 是体检报告上的指标，是一种远离医院的生活状态，还是一个能随时解答焦虑AI医生。 如今的大健康行业，所有参与者都在试图重新定义这个问题的答案。 在中国，20万亿元级的大健康市场正处于数字化、智能化转型的关键期。不仅有蚂蚁、京东等互联网企业布局，还有平安好医生、北电数智等企业入场。 这是需求推动的结果。过去几十年来，我国建立起全球最为庞大的医疗系统和医保体系。但随着我国人口老龄化加剧，看病难、看病贵等问题带来的供需 缺口进一步加大，也推动大健康行业进入新的阶段。 曾经跑马圈地的流量争夺赛早已落幕，一场围绕用户心智的精准角逐已然开启。一个确定性的行业共识时，行业未来的领跑者，是具备医疗生态整合能力 的全能型选手，即能利用技术和资源为患者提供普惠的医疗 ...

编者按： 岁末将至，站在2025年的时间节点回望，技术浪潮的奔涌、消费需求的变迁、商业模式的迭代，构成了全新的商业图景。连线Insight推出年终 盘点专题系列，试图捕捉不同企业在这幅变局图景中，如何应对挑战、抓住机遇。本期为第一篇，关注大健康行业。 当人们谈论"健康"，人们究竟在谈论什么？ 是体检报告上的指标，是一种远离医院的生活状态，还是一个能随时解答焦虑AI医生。 如今的大健康行业，所有参与者都在试图重新定义这个问题的答案。 在中国，20万亿元级的大健康市场正处于数字化、智能化转型的关键期。不仅有蚂蚁、京东等互联网企业布局，还有平安好医生、北电数智等企业入场。 这是需求推动的结果。过去几十年来，我国建立起全球最为庞大的医疗系统和医保体系。但随着我国人口老龄化加剧，看病难、看病贵等问题带来的供需 缺口进一步加大，也推动大健康行业进入新的阶段。 曾经跑马圈地的流量争夺赛早已落幕，一场围绕用户心智的精准角逐已然开启。一个确定性的行业共识时，行业未来的领跑者，是具备医疗生态整合能力 的全能型选手，即能利用技术和资源为患者提供普惠的医疗解决方案。 这背后，医疗大模型的规模化落地成为关键引擎，大模型赋能千行百业，百 ...

Group 1 - The core viewpoint of the articles emphasizes the transformative potential of AI in the healthcare sector, moving from concept to practical application, with a focus on enhancing efficiency and precision in medical services [1][2][6] - The healthcare industry is undergoing a significant shift from "experience-based decision-making" to "algorithm-driven empowerment," highlighting the necessity for high-quality data and specialized models to improve diagnostic accuracy and reduce R&D costs [1][2] - The integration of AI capabilities into grassroots healthcare and extending services beyond hospitals is seen as a crucial strategy to address uneven distribution of medical resources and create new growth opportunities within the industry [1][2] Group 2 - North Electric Intelligence (北电数智) is positioning itself as a key player in the AI healthcare landscape, showcasing its strategic layout and practical achievements in building a comprehensive intelligent healthcare system during the recent symposium [3] - The company is developing a dual-driven system of "general data + specialized disease data" to ensure compliant data circulation, which supports high-quality data for model training and facilitates the upgrade of AI from general assistance to specialized expertise [3] - North Electric Intelligence has successfully implemented its intelligent healthcare system at the China-Japan Friendship Hospital, demonstrating significant improvements in efficiency, such as a 20% reduction in diagnosis time and a 75% increase in medical record writing efficiency [4][6]

当前，睡眠不足、质量不好已成为人们健康的"隐形杀手"。据中国睡眠研究会今年3月发布的《2025年 中国睡眠健康调查报告》显示，中国18岁及以上人群睡眠困扰率约为48.5%，超3亿人存在睡眠障碍。 长期睡眠不足或睡眠质量差，不仅会导致注意力下降、情绪紊乱，影响生活质量，更与高血压、糖尿 病、心脑血管疾病、抑郁症等多种慢性病的发生发展密切相关。 随着北电数智与北京清华长庚医院在睡眠大模型领域合作的深入开展，有望推动将专家级的睡眠诊疗经 验赋能至基层，提升基层医疗服务效率与智能化水平，推动医疗普惠进程。 由于睡眠障碍的病因复杂，归因判断高度依赖临床经验。三甲医院睡眠医学科的专业医师，主要通过专 业问诊形式对患者的睡眠障碍进行分类与诊断。相较而言，基层医疗机构普遍面临睡眠专科资源匮乏、 诊断能力不足的困境，难以精准把握问诊重点、科学判断病因关联，导致大量患者难以在早期获得精准 筛查，往往延误干预时机。 近年来，人工智能技术的快速突破为破解这一难题提供了新路径，通过学习并融合专家经验构建专病大 模型，有望将优质诊疗能力赋能至基层。基于这一探索方向，10月16日，北京清华长庚医院与北京电子 数智科技有限责任公司(简称"北 ...

Core Insights - Artificial Intelligence (AI) is becoming a core engine driving global industrial transformation, but faces significant challenges in the medical field, including difficulties in commercializing domestic computing power, applying AI in real-world scenarios, and releasing data value [1][2] Group 1: AI in Healthcare - The collaboration between Beidian Zhizhi and the Japan-China Friendship Hospital offers a new approach to overcoming challenges in AI healthcare development, serving as a successful example of how AI can empower traditional industries [1] - The Chinese government has emphasized the integration of AI in healthcare, issuing policies to promote the use of AI technologies to innovate medical service models and improve efficiency and quality [1] Group 2: Challenges in AI Implementation - The commercialization of domestic computing power is hindered by high infrastructure costs, fragmented market demand, and immature business models, making it difficult for medical institutions to leverage advanced computing power [2] - The medical industry's professional and regulatory nature requires extensive clinical trials for AI technologies, which often fail to meet strict regulatory standards, complicating their clinical application [2] - The release of data value is challenged by the fragmentation and lack of standardization in medical data, as well as legal and technical issues surrounding patient privacy and data sharing [2] Group 3: Solutions and Innovations - Beidian Zhizhi's "Spark Medical Base" is a key solution for addressing these challenges, providing a one-stop empowerment system for medical institutions from foundational technology to application development [4] - The "Zhongri Sakura Agent Development Platform" developed in collaboration with the Japan-China Friendship Hospital integrates DeepSeek-R1, enabling customized development that aligns with hospital workflows and enhances clinical efficiency [5] - The establishment of a trusted data application platform allows for the integration and cleaning of hospital data, ensuring security and privacy, which facilitates the release of medical data value [6] Group 4: Impact and Future Directions - The AI solutions implemented at the Japan-China Friendship Hospital have shown significant results, including a 20% reduction in diagnosis time, a 15% decrease in misdiagnosis rates, and a 75% increase in medical record writing efficiency [6] - The AI pharmaceutical market is projected to reach $2.994 billion by 2026, with AI technologies reshaping drug innovation processes and expanding into personalized medicine and rare disease drug development [7] - Future collaborations aim to explore more applications in clinical decision support, patient services, and resource management, contributing to the intelligent transformation of the healthcare industry [9]