Preventive Medicine Intelligence · Hangzhou · Est. 2025 预防医疗智能 · 杭州 · 成立于2025年
📰 已获 36氪 正式报道 → Featured on 36Kr →

The Infrastructure
for Preventive
Medicine 构建预防医疗
决策基础设施

ReHealth AI converts long-term health behaviors into measurable, quantifiable prevention outcomes — enabling healthcare institutions and payers to act before disease strikes. 睿禾健康将长期健康行为转化为可量化的预防医疗证据——让医疗机构与支付方在疾病发生前有效行动。

Explore the Platform了解平台 Request API Access申请 API 接入
The Problem核心痛点

Prevention has value.
It just can't be billed.预防有价值。
但无法被结算。

Over 70% of cardiovascular costs occur when intervention could have happened 3–5 years earlier. The problem isn't lack of technology — it's the inability to prove prevention worked. 心脑血管疾病70%以上的医疗支出，发生在本可提前3-5年干预的阶段。问题不是缺乏技术——而是无法证明预防发挥了作用。

"Prevention isn't without value. Prevention outcomes just can't be settled." 「不是预防没价值，而是预防结果无法被结算。」
📱
Health Apps健康管理 App
Track behavior but cannot prove clinical outcomes. Payers don't accept them.能记录行为，但无法证明临床效果，支付方不认可。
🤖
Medical AI Platforms医疗 AI 平台
Only predict risk. Don't manage, intervene, or attribute outcomes.只预测风险，不管理、不干预、不对结果归因。
⌚
Wearable Devices可穿戴设备
Collect data continuously but zero causal analysis.有连续数据，但零因果分析，无法产生预防证据。
💬
General AI通用 AI
Non-compliant, non-auditable for clinical environments.不合规、不可审计，无法用于临床环境。
Our Solution解决方案

From behavior to billable evidence从行为到可结算的医疗证据

ReHealth Core runs the complete prevention loop — prediction to settlement — starting with cardiovascular disease. ReHealth Core 跑通从预测到结算的完整预防闭环，聚焦心脑血管疾病。

01
Risk Trajectory Prediction风险轨迹预测
Multimodal time-series modeling anticipates risk 3–5 years ahead多模态时序建模，提前3-5年预测心脑血管风险
02
Personalized Intervention个性化干预
AI-generated plans adapted to individual health profiles基于个体健康档案生成个性化干预方案
03
Causal Attribution因果归因分析
PSM and DID verify whether interventions genuinely changed riskPSM、DID验证干预是否真正改变了风险轨迹
04
Prevention Settlement预防效果结算
Causal evidence enables insurers to pay for prevention outcomes因果证据让保险机构可以为预防效果付费
rehealth_core_api.py
# ReHealth Core — Prevention Intelligence API

from rehealth import CoreAPI

# Initialize with federated learning
api = CoreAPI(mode="federated")

# Risk trajectory prediction
result = api.predict_risk({
  "patient_id": "P_2847",
  "timeframe_years": 5,
  "focus": "cardiovascular"
})

>>> result.risk_score: 0.82
>>> result.intervention: "high_priority"
>>> result.causal_evidence: "PSM_validated"
>>> result.settlement_ready: True
Products产品

Two products. One infrastructure.两款产品。一套基础设施。

ReHealth Core powers institutions. BodyUP engages individuals. Together they form a closed-loop prevention ecosystem.ReHealth Core 赋能机构，BodyUP 连接个人，共同构成闭环预防生态。

B2B · Enterprise
⚕️
ReHealth Core
AI decision infrastructure for healthcare institutions, insurers, and enterprise health面向医疗机构、保险公司、企业健康管理的AI决策基础设施
  • ✓Federated learning cardiovascular risk API — data never leaves your institution联邦学习心脑血管风险API——数据不出院
  • ✓Long-term health memory modeling for individual risk trajectory tracking长期健康记忆建模，追踪个体风险演进轨迹
  • ✓Causal attribution analysis to verify real intervention effectiveness因果归因分析，验证干预措施的真实有效性
  • ✓Prevention outcome settlement interface for insurers and payers预防效果结算接口，服务保险公司与支付方
Serving服务对象
Hospitals · Insurers · Enterprise Health医疗机构 · 保险公司 · 企业健康管理
Consumer · C2B
💪
BodyUP
Your AI health companion — collecting data, detecting early signals, guiding proactive actionAI孪生健康伙伴——收集数据、识别早期风险信号、引导主动健康行动
  • ✓AI digital twin with long-term health memory具备长期健康记忆的AI数字孪生
  • ✓Gamified health missions with continuous personalized feedback游戏化健康任务，持续个性化反馈
  • ✓Multi-source physiological and behavioral data aggregation多源生理与行为数据聚合分析
  • ✓Early health awareness signals — not diagnoses, informed perspective早期健康意识信号——提升健康认知，非医疗诊断
Available On平台覆盖
WeChat Mini Program ✓ iOS · Coming研发中 Android · Coming研发中 HarmonyOS · Coming研发中
Technology核心技术

Built for clinical trust.为临床信任而构建。

Three technical moats that existing solutions cannot replicate — built from the ground up for clinical and regulatory constraints. 三个现有方案无法复制的技术壁垒——从底层为临床和监管环境的约束而设计。

🧠
Temporal Health Memory Architecture时序健康记忆架构
Not conversation memory. Not RAG. True longitudinal health trajectory modeling that builds individual risk profiles across months and years of behavioral and physiological data.不是对话记忆，不是RAG，而是真正的纵向健康轨迹建模。跨越数月、数年的行为与生理数据，构建个体风险演进档案。
🔒
Medical-Grade Federated Learning医疗级联邦学习
Patient data never leaves the institution. Our federated model continuously improves across hospital networks while fully satisfying data compliance. Bias correction built in.患者数据始终留在机构本地。联邦模型跨医院网络持续进化，满足数据合规要求，内置偏见校正流程。
⚖️
Causal Attribution as Core Objective因果归因为核心目标
Using PSM as the core method, progressively introducing DID and synthetic controls to build causal evidence acceptable to clinical standards and payers — not just prediction accuracy.以PSM为核心方法，逐步引入DID和合成对照，构建可被临床标准和支付方接受的因果证据体系。
Open Source开源项目

We build in public.我们公开构建。

ReHealth AI contributes to the open-source community. Our multi-agent health simulation engine is freely available under Apache 2.0. ReHealth AI 积极回馈开源社区。我们的多智能体健康干预模拟引擎已以 Apache 2.0 协议开源。

⚙️ ReHealthAI-HealthAgent Open Source

A multi-agent simulation framework modeling how human emotion, behavior, and physiology interact during health interventions. Predict patient trajectories, test behavioral protocols, and evaluate compliance strategies — without real-world trials. Built for AI researchers, digital health builders, and multi-agent experimentation. 多智能体健康干预模拟框架，建模人类情绪、行为与生理指标的交互。无需真实临床试验，即可预测患者轨迹、测试干预方案、评估依从性策略。专为AI研究者和数字健康开发者构建。

Python Multi-Agent AI Health Simulation Behavioral Modeling Apache 2.0
View on GitHubGitHub 查看
Apache 2.0 · Free for commercial use允许商业使用
Multi-Agent · Emotion, Compliance, Physiology, Intervention agents情绪、依从、生理、干预多智能体
Python · DeepSeek API
API AccessAPI 接入

Build with ReHealth AI.
By invitation only.基于 ReHealth AI 构建。
定向邀请接入。

ReHealth Core API is available to qualified healthcare institutions, insurance companies, and enterprise health management partners. We review each application personally. ReHealth Core API 向符合资质的医疗机构、保险公司和企业健康管理合作伙伴开放。我们亲自审核每一份申请。

📊
Risk Stratification API风险分层 API
Cardiovascular high-risk identification with federated learning — data stays in your institution联邦学习心脑血管高风险识别——数据不出院
🔗
HIS System IntegrationHIS 系统集成
Deep integration with existing hospital information systems与现有医院信息系统深度集成
⚖️
Prevention Settlement Interface预防结算接口
Quantified causal evidence for insurers to pay for prevention outcomes量化因果证据，支持保险机构为预防效果付费
🛡️
Compliance Guaranteed合规保障
Federated architecture — patient data never leaves your institution联邦架构，患者数据不出院，内置数据合规机制

Apply for API Access申请 API 接入

We review all applications and respond within 3 business days.我们审核每份申请，3个工作日内回复。

We do not share your information with third parties.我们不会向第三方共享您的信息。

Get in Touch联系我们

Let's build the future of prevention together.携手共建预防医疗的未来。

Whether you're a healthcare institution, insurer, enterprise, researcher, or partner — we'd love to hear from you.无论您是医疗机构、保险公司、企业、研究机构还是合作伙伴——我们期待与您交流。

Apply for API申请 API peijun@rehealth-ai.cn