medical-writing
临床文档撰写自动化
面向 CSR/protocol/lay summary 的可追溯自动生成。
2 家 vendor · 2 个产品 · 1 篇文献
场景覆盖率评分: 28/100
prototype 18
临床 AI Vendor 情报雷达
按应用场景组织 CRO 与临床科技版图：谁在解决什么问题、使用什么技术、哪些文献正在推动前沿。
更新时间: 3/1/2026, 12:00:00 AM · signals: 7/21/2026, 11:32:57 PM
site-feasibility
站点可行性与启动
站点排名、启动风险预测与启动加速。
2 家 vendor · 2 个产品 · 0 篇文献
场景覆盖率评分: 24/100
patient-recruitment
患者招募智能
可入组人群测算、招募预测与渠道优化。
5 家 vendor · 5 个产品 · 1 篇文献
场景覆盖率评分: 54/100
protocol-intelligence
Protocol 智能
协议解析、入排判断与规则证据检索。
4 家 vendor · 4 个产品 · 1 篇文献
场景覆盖率评分: 48/100
data-quality-rbqm
RBQM 与数据质量
风险导向监查、异常检测与数据问题分流。
3 家 vendor · 3 个产品 · 1 篇文献
场景覆盖率评分: 35/100
synthetic-control
数字孪生与合成对照
历史对照构建与样本量优化仿真。
3 家 vendor · 3 个产品 · 1 篇文献
场景覆盖率评分: 35/100
supply-rtsm
RTSM 与供应优化
药物供应预测、仓配平衡与随机化运营。
0 家 vendor · 0 个产品 · 0 篇文献
场景覆盖率评分: 0/100
safety-phv
安全与药物警戒信号
安全病例分流、信号优先级与叙事自动化。
1 家 vendor · 1 个产品 · 0 篇文献
场景覆盖率评分: 11/100
submission-closeout
递交与关库加速
注册包就绪度、资料拼装与关库编排。
3 家 vendor · 3 个产品 · 1 篇文献
场景覆盖率评分: 37/100
portfolio-control-tower
组合管理控制塔
里程碑、预算与 FTE 的统一控制与情景规划。
4 家 vendor · 4 个产品 · 0 篇文献
场景覆盖率评分: 44/100
multimodal-assessment
影像与多模态评估
以计算机视觉支持影像终点评估与多模态结局分析。
1 家 vendor · 1 个产品 · 1 篇文献
场景覆盖率评分: 19/100
Vendor 解决方案
IQVIA (CRO)
AI-enabled Trial Optimization Suite
livePortfolio-level optimization and operational forecasting for trial execution.
站点可行性与启动 · 组合管理控制塔 · 患者招募智能
Machine Learning · Optimization · Bayesian Modeling
打开Parexel (CRO)
Clinical Data Intelligence Services
liveRisk analytics and data quality insights for study teams.
RBQM 与数据质量 · 递交与关库加速
Machine Learning · NLP / IE
打开Medidata (Platform)
Medidata AI Modules
liveAI capabilities across study build, risk monitoring, and workflow support.
站点可行性与启动 · RBQM 与数据质量 · Protocol 智能
Machine Learning · NLP / IE · Agent Systems
打开Veeva (Platform)
Vault Clinical + AI Search/Assist
betaDocument lifecycle and regulatory workflows with AI-assisted retrieval and drafting.
临床文档撰写自动化 · 递交与关库加速 · Protocol 智能
GenAI · Knowledge Graph · Rules Engine
打开Saama (Clinical AI)
Clinical Analytics Cloud
liveAutomated data review and quality risk surfacing.
RBQM 与数据质量 · 组合管理控制塔
Machine Learning · Optimization · Continuous Learning
打开Flatiron Health (RWD)
Oncology RWE Data Products
liveReal-world evidence datasets for trial planning and endpoint contextualization.
患者招募智能 · 数字孪生与合成对照
Causal Inference · Machine Learning · Bayesian Modeling
打开PathAI (Clinical AI)
Digital Pathology AI
liveAI-assisted pathology quantification and endpoint support.
影像与多模态评估
Computer Vision · Self-supervised Learning
打开Tempus (Clinical AI)
Precision Medicine AI Platform
liveMolecular + clinical data insights for cohort strategy and trial matching.
患者招募智能 · Protocol 智能
Machine Learning · Knowledge Graph · NLP / IE
打开Insilico Medicine (Biotech AI)
Generative Discovery/Design Stack
liveAI-first drug discovery platform with design and optimization loops.
组合管理控制塔
GenAI · Reinforcement Learning · Continuous Learning
打开Unlearn (Clinical AI)
Digital Twin Generator
betaSynthetic control modeling to improve trial efficiency.
数字孪生与合成对照 · 患者招募智能
Simulation / Digital Twin · Bayesian Modeling · Causal Inference
打开Datavant (Data Infrastructure)
Privacy-preserving Data Linkage
liveSecure linkage infrastructure across healthcare datasets.
患者招募智能 · 安全与药物警戒信号 · 数字孪生与合成对照
Machine Learning · Optimization
打开OpenAI / Microsoft Ecosystem (Foundation Model)
Enterprise AI Copilot Stack
liveGeneral-purpose language and agent systems adaptable to clinical workflows.
临床文档撰写自动化 · Protocol 智能 · 递交与关库加速 · 组合管理控制塔
GenAI · Agent Systems · NLP / IE
打开技术面
Agent SystemsBayesian ModelingCausal InferenceComputer VisionContinuous LearningGenAIKnowledge GraphMachine LearningNLP / IEOptimizationReinforcement LearningRules EngineSelf-supervised LearningSimulation / Digital Twin
Vendor 对比矩阵
| Vendor | 场景支持 | GenAI | Agent | CV | 仿真 | 规则 | 成熟度 |
|---|---|---|---|---|---|---|---|
| IQVIA | 是 | - | - | - | - | - | 100% |
| Parexel | 是 | - | - | - | - | - | 100% |
| Medidata | 是 | - | 是 | - | - | - | 100% |
| Veeva | 是 | 是 | - | - | - | 是 | 60% |
| Saama | 是 | - | - | - | - | - | 100% |
| Flatiron Health | 是 | - | - | - | - | - | 100% |
| PathAI | 是 | - | - | 是 | - | - | 100% |
| Tempus | 是 | - | - | - | - | - | 100% |
| Insilico Medicine | 是 | 是 | - | - | - | - | 100% |
| Unlearn | 是 | - | - | - | 是 | - | 60% |
| Datavant | 是 | - | - | - | - | - | 100% |
| OpenAI / Microsoft Ecosystem | 是 | 是 | 是 | - | - | - | 100% |
文献信号
Patient simulation frameworks for virtual trial cohorts
2025 · Journal/Preprint Monitoring · high
Simulation / Digital Twin · Bayesian Modeling
打开Risk-based quality monitoring using anomaly detection in multicenter trials
2024 · Journal/Conference Monitoring · high
Machine Learning · Continuous Learning
打开LLM-assisted clinical writing with source traceability constraints
2025 · Preprint Monitoring · medium
GenAI · Rules Engine
打开Multimodal endpoint modeling across imaging and tabular clinical signals
2025 · Conference Monitoring · medium
Computer Vision · Machine Learning · Causal Inference
打开最新自动采集信号
Rater State Bias in RLHF Preference Data: An Audit Framework
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开Design and Validation of a Lightweight 1D CNN for Affective Touch Classification in Soft Plush Companions
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开Some Large Language Models Exhibit Consistent Risk Attitudes
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开A Survey on GNN-based Link Prediction: Techniques, Applications, and Challenges
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开PlanFlip: Attacking Multi-Agent LLM Systems via Planning-Phase Prompt Injection
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开Deterministic Replay for AI Agent Systems
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开Generative Ontology Induction: Domain-Agnostic Schema Discovery from Document Corpora Using Large Language Models
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment
2026-07-21 · arXiv cs.AI · literature
Machine Learning
自动归类置信度: 45%
打开Emergent topological structure in spontaneous brain-organoid activity
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开The Origins of Transient Bimodality
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开Orientation Reading by Production Vision-Language Models on Optotype Charts: A Controlled Multi-Model Evaluation Across Reasoning Modes, Prompts, and Access Modalities
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开The Positive Experience Principle: Forecasting Conscious Choices with AI Embeddings
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开Evolutionary dynamics in public goods games with general frequency-dependent returns
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开Laboratory Trajectories Improve Kidney Failure Risk Estimation
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开Post-transcriptional Regulation of Stochastic Gene Expression Conditioned on Large Deviations
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开Network Characteristics of Individual Pigments in Cyanobacterial Photosystem II Core Complexes
2026-07-21 · arXiv q-bio · literature
Machine Learning
自动归类置信度: 45%
打开update log
Prototype Change Log
2026-03-01 · v0.2.0
覆盖率评分 + Vendor 对比矩阵 + 归类置信度层
- - 新增场景覆盖率评分（0-100），综合 vendor/product/literature/technology 覆盖面。
- - 新增 vendor 对比矩阵（场景支持、GenAI、Agent、CV、仿真、规则、成熟度）。
- - 新增自动采集信号的归类置信度（基于关键词命中密度）。
2026-03-01 · v0.1.0
Vendor Intelligence 雷达基线版本
- - 新增场景优先卡片，覆盖不局限于 GenAI 的临床 AI 应用场景。
- - 建立 vendor/product 与场景、技术模态（CV、RL、仿真、因果、Agent）映射。
- - 新增文献信号面板与每日自动采集信号面板。
- - 新增每日同步脚本：RSS/Atom + 关键词映射到场景与技术。