OpenLens AI - Tsinghua University's AI Assistant for Medical Research

What is OpenLens AI?

OpenLens AI, developed by the Department of Automation at Tsinghua University, is a fully autonomous AI research assistant designed specifically for medical research. Through multi-agent collaboration, it automates the entire process from literature review, experimental design, data analysis to paper generation, significantly improving the efficiency of medical research and reducing the research cycle from months to hours. The tool integrates multiple quality control mechanisms to ensure the reliability and reproducibility of research results. The emergence of OpenLens AI marks the beginning of a "zero-human" era in medical research, bringing about a major transformation in the research paradigm.

Main functions of OpenLens AI

• literature reviewIt automatically retrieves and synthesizes relevant literature, using a ReAct-based reasoning framework to provide a comprehensive and accurate theoretical foundation for research.
• Experimental DesignIt generates detailed experimental plans based on the user's research questions, converts them into executable code, and supports various experimental design requirements.
• Data AnalysisOpenLens AI processes various medical data (such as time series and genomic information) to generate structured, interpretable insights, while providing visualizations, statistical summaries, and natural language interpretations.
• Thesis writingIt integrates the output of all modules to generate high-quality scientific papers, supports LaTeX format, and ensures the quality and format consistency of figures and tables.
• Quality controlIt integrates mechanisms such as academic rigor checks, evidence traceability checks, literature checks, and visual language feedback to ensure the reliability and reproducibility of research results.

The technical principles of OpenLens AI

• Multi-agent collaboration:
  • Supervisor ModuleAs the global coordinator, it breaks down user queries into structured subtasks, ensuring transparency and interpretability throughout the research process.
  • Literature reviewer: Construct an autonomous knowledge exploration pipeline and use a ReAct-based reasoning framework to retrieve and synthesize relevant literature.
  • Data AnalystCoordinate multi-stage data processing workflows to transform raw medical data into structured, interpretable insights.
  • encoder: Convert advanced experimental plans into executable code and evaluate the output quality using a visual language model.
  • LaTeX writersIt integrates the outputs of all preceding modules to generate a publication-quality scientific paper.
• Modular architecture:Each agent focuses on a specific task, and modular design enhances the system's flexibility and scalability.Intelligent agents collaborate through standardized interfaces, ensuring the continuity and efficiency of the research process.
• Quality control mechanismThe system automatically verifies the rationality of experimental methods, detects common pitfalls such as data leakage and inappropriate performance metrics, and ensures the scientific rigor of research results. It links each research claim to underlying evidence, generating structured traceability reports to ensure research transparency. The system verifies all cited references, confirms the accuracy of metadata, removes unverifiable entries, and guarantees the reliability of academic background. At critical stages, the system evaluates the quality of graphs and charts, providing perceptual feedback to enhance the readability and scientific validity of results.

OpenLens AI project address

• Project official websitehttps://openlens.icu/
• GitHub repositoryhttps://github.com/jarrycyx/openlens-ai
• arXiv technical paper: https://arxiv.org/pdf/2509.14778

Application scenarios of OpenLens AI

• Disease prediction modeling: Develop disease prediction models based on historical patient data, such as the accuracy of predicting acute kidney injury (AKI).
• Drug interaction studiesAnalyzing the latest findings on drug interactions helps improve medication safety and efficacy.
• Causal relationship analysisUsing structural causal models, we explore the potential causal drivers of complex medical problems, such as the reasons for prolonged ICU stays.
• Medical data analysisProcess and analyze medical datasets (such as MIMIC-IV and eICU datasets) to generate statistical reports and visualizations.
• Scientific paper generationThe entire process, from researching questions to writing papers, is automated, generating high-quality scientific papers and accelerating the medical research process.