医疗人工智能公司DeHaze筹集320万欧元种子轮融资，用于构建首个用于慢性病检测的基础人工智能模型-动脉网

Chronic diseases cost $8 trillion annually: A Munich-based startup aims to change that

D-Pharm 等信源发布 2026-05-06 12:16



可切换为仅中文







dehaze, the Munich-based

去雾，总部位于慕尼黑的

healthcare

医疗保健

AI company, announced the closing of a €3.2 million seed round led by YZR Capital and DN Capital, with participation from Angel Invest, ZOHO and Better Ventures.

人工智能公司宣布完成了由YZR Capital和DN Capital领投、Angel Invest、ZOHO和Better Ventures参与的320万欧元种子轮融资。

The investment confirms dehaze’s commercial traction with international payers and its position as one of the most promising players to build the first foundational AI model for chronic disease detection. The company’s mission is to build its causal AI model that detects and explains the risks of chronic diseases so that patients can be treated sooner and more effectively, and live healthier lives..

这项投资证实了Dehaze在国际支付方中的商业吸引力，以及其作为最有前途的构建首个慢性病检测基础人工智能模型的公司之一的地位。该公司的使命是构建其因果人工智能模型，以检测和解释慢性病的风险，从而使患者能够更早、更有效地接受治疗，并过上更健康的生活。

According to WHO, chronic diseases are the single largest cost driver in every healthcare system on earth, responsible for roughly seven out of every ten deaths globally and more than $8 trillion in annual spend. Physicians have access to enormous volumes of health data, but the time and tools to review less than 3% of it before making a clinical decision.

根据世界卫生组织的数据，慢性疾病是全球每个医疗保健系统中最大的单一成本驱动因素，大约占全球死亡人数的七成，并导致每年超过8万亿美元的支出。医生可以访问大量的健康数据，但在做出临床决策之前，他们有时间和工具来审查的数据不到3%。

As a result, more than 31% of chronic diseases are overlooked at the point where they could still be prevented, slowed, or treated at a fraction of the later cost..

因此，超过 31% 的慢性病在本可以预防、减缓或以更低的成本进行治疗的阶段被忽视了。

Artificial Intelligence for faster and more reliable radiological findings

人工智能助力更快、更可靠的放射学诊断结果

Read news

读新闻

dehaze is the first company to close this gap. Its foundational model processes the full breadth of available patient data to identify individuals at elevated risk of chronic disease continuously, at population scale, and with causal explanations clinicians and payers can act on. Unlike general-purpose LLMs, which are structurally unable to draw causal conclusions from sparse and heterogeneous medical data, dehaze is built from first principles for the realities of healthcare data.

Dehaze 是第一家填补这一空白的公司。其基础模型处理所有可用的患者数据，以持续识别慢性病高风险个体，在人群规模上进行，并提供临床医生和支付者可以采取行动的因果解释。与通用的大型语言模型（LLM）不同，这些模型在结构上无法从稀疏且异质的医疗数据中得出因果结论，而 Dehaze 则是从头开始构建，以适应医疗数据的现实情况。

The result: payers can lower medical loss ratios by up to 10%, and patients can be treated sooner and more effectively..

结果：支付者可以将医疗损失率降低多达10%，患者可以更快、更有效地得到治疗。

dehaze’s engineering, research and product teams are headquartered in Munich, with deep roots in German and European deep-tech research, supported by multiple public grants from the German Federal Ministry of Research, Technology and Space and the German Federal Ministry of Economic Affairs and Energy.

dehaze的工程、研究和产品团队总部位于慕尼黑，深深扎根于德国和欧洲的深科技研究，并获得了来自德国联邦教育与研究部、德国联邦经济事务和能源部的多项公共资助。

Its commercial footprint is already global, with a growing pipeline of payers actively onboarding the platform, set to be enhanced by new modules focusing on recommending the next best action and full traceability. The company plans to expand the core team with medical and technical expertise to enhance the technological lead and the commercial team to meet growing commercial demand..

其商业版图已遍布全球，目前不断有支付方加入该平台，未来还将通过新增模块得到进一步提升，这些模块专注于推荐下一步最佳行动以及实现全程可追溯性。公司计划扩充核心团队，引入医疗和技术专家以巩固技术领先地位，同时扩大商业团队以满足不断增长的商业需求。

Marius Klages, Co-Founder and CEO of dehaze, commented: “Chronic disease is the biggest and most expensive problem in healthcare, and it has been unsolved for one simple reason: the data exists, but no one has been able to use it. Doctors see less than 3% of what’s available before making a decision, and as a result roughly a third of chronic diseases go undetected until it’s too late or too expensive to act.

德哈兹公司的联合创始人兼首席执行官马里乌斯·克拉格斯评论道：“慢性病是医疗保健领域最大且最昂贵的问题，它一直未能得到解决的原因很简单：数据是存在的，但没有人能够利用它。医生在做决策前只能看到不到3%的可用信息，因此大约三分之一的慢性病在被发现时已经为时已晚或治疗成本过高。”

That is the gap we close. We built dehaze in Munich, from first principles, as a foundational AI model for chronic disease detection - not a chatbot, not a dashboard. Our customers are global from day one, because the problem is global from day one. The speed at which payers are signing with us confirms what we believed when we started: this is a category that will be defined over the next few years, and dehaze is going to define it.”.

这就是我们所填补的空白。我们在慕尼黑从头开始构建了dehaze，作为一个用于慢性病检测的基础AI模型——不是聊天机器人，也不是仪表盘。我们的客户从第一天起就是全球化的，因为这个问题从一开始就是全球性的。支付方与我们签约的速度证实了我们创业时的信念：这一领域将在未来几年内被定义，而dehaze将会定义它。

Markus Feuerecker, Co-Founder & General Partner at YZR Capital: “At YZR Capital we focus on identifying global multi-billion USD opportunities within healthcare, that can be tackled with cutting edge AI. Oftentimes by doing that, complete value chains are innovated. With dehaze this is happening at a complete system level, finally utilizing the masses of unstructured data available in care systems for effective chronic disease prevention through causal AI.

马库斯·菲勒克，YZR资本的联合创始人兼普通合伙人：“在YZR资本，我们专注于识别医疗保健领域内可借助尖端人工智能解决的全球性数十亿美元的机会。很多时候，通过这样做，整个价值链都得到了创新。对于dehaze来说，这一创新正在全面系统层面发生，最终利用护理系统中大量可用的非结构化数据，通过因果人工智能实现有效的慢性病预防。”

Quite naturally this is one of the most challenging tasks within health AI, therefore requiring unique teams that combine a deep technological foundation with a profound understanding of the needs of the global health insurance clients – in dehaze we found such a team. For that reason, we are particularly excited to partner with them on their mission to become the globally leading AI platform for chronic disease detection.”.

很自然，这是健康人工智能领域中最艰巨的任务之一，因此需要独特的团队，将深厚的技术基础与对全球健康保险客户需求的深刻理解相结合——在德哈兹，我们找到了这样的团队。出于这个原因，我们特别兴奋能够与他们合作，助力他们成为全球领先的慢性病检测人工智能平台。

Gülsah Wilke, Partner and Head of the German Office at DN Capital: “dehaze is not an LLM for medicine. Marius and the team are building the unglamorous but vitally important and scientifically rigorous layer that global healthcare has been missing; a foundation model designed for patient data, by people who understand patient data, that will shift the system from reactive to preventive.

古尔萨·威尔克，DN Capital的合伙人兼德国办事处负责人：“dehaze并非医学领域的大型语言模型。马里乌斯及其团队正在构建全球医疗保健领域一直缺失的、虽不引人注目但却至关重要且科学严谨的基础层；这是一个由理解患者数据的人设计的、专为患者数据打造的基础模型，它将使医疗系统从被动反应转变为预防为主。”

dehaze is the kind of deep-tech, mission-driven company Germany should be famous for championing.”.

“去雾公司正是德国应该鼎力支持的那种深科技、使命驱动型公司。”

https://www.bionity.com/en/news/1188632/chronic-diseases-cost-8-trillion-annually-a-munich-based-startup-aims-to-change-that.html

https://www.bionity.com/zh/news/1188632/慢性病每年花费8万亿美元-慕尼黑一家初创公司旨在改变这一现状.html

17 Million for Future Cancer Diagnoses

1700万用于未来的癌症诊断

Read news

读新闻