![[2602.19158] DoAtlas-1: A Causal Compilation Paradigm for Clinical AI](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.19158] DoAtlas-1: A Causal Compilation Paradigm for Clinical AI
Summary
The paper presents DoAtlas-1, a novel causal compilation paradigm for clinical AI that transforms medical evidence into executable code, enhancing clinical auditability and causal reasoning.
Why It Matters
This research addresses critical limitations in current medical foundation models, which often fail to quantify intervention effects or validate claims. By introducing a structured approach to causal reasoning, DoAtlas-1 aims to improve the reliability and transparency of clinical AI applications, which is essential for advancing healthcare technology.
Key Takeaways
- DoAtlas-1 standardizes medical evidence into structured estimand objects.
- The paradigm supports six executable causal queries for enhanced analysis.
- Achieved 98.5% canonicalization accuracy and 80.5% query executability.
- Shifts focus from text generation to auditable and verifiable causal reasoning.
- Promotes better clinical decision-making through improved evidence handling.
Computer Science > Artificial Intelligence arXiv:2602.19158 (cs) [Submitted on 22 Feb 2026] Title:DoAtlas-1: A Causal Compilation Paradigm for Clinical AI Authors:Yulong Li, Jianxu Chen, Xiwei Liu, Chuanyue Suo, Rong Xia, Zhixiang Lu, Yichen Li, Xinlin Zhuang, Niranjana Arun Menon, Yutong Xie, Eran Segal, Imran Razzak View a PDF of the paper titled DoAtlas-1: A Causal Compilation Paradigm for Clinical AI, by Yulong Li and 11 other authors View PDF HTML (experimental) Abstract:Medical foundation models generate narrative explanations but cannot quantify intervention effects, detect evidence conflicts, or validate literature claims, limiting clinical auditability. We propose causal compilation, a paradigm that transforms medical evidence from narrative text into executable code. The paradigm standardizes heterogeneous research evidence into structured estimand objects, each explicitly specifying intervention contrast, effect scale, time horizon, and target population, supporting six executable causal queries: do-calculus, counterfactual reasoning, temporal trajectories, heterogeneous effects, mechanistic decomposition, and joint interventions. We instantiate this paradigm in DoAtlas-1, compiling 1,445 effect kernels from 754 studies through effect standardization, conflict-aware graph construction, and real-world validation (Human Phenotype Project, 10,000 participants). The system achieves 98.5% canonicalization accuracy and 80.5% query executability. This paradigm shifts m...
