基于真实世界数据的高保真纵向患者模拟
人工智能
2026-01-27 v1 机器学习
摘要
模拟是探索不确定性的强大工具，其在临床医学中的潜力是变革性的，包括个性化治疗规划和虚拟临床试验。然而，由于受到复杂的生物和社会文化影响，模拟患者轨迹具有挑战。本文表明，真实临床记录可以被用来经验性地对患者时间线进行建模。我们开发了一个生成式模拟模型，该模型以患者历史为输入，并综合出细粒度、逼真的未来轨迹。该模型在超过2亿条临床记录上进行预训练。它产生了高保真的未来时间线， closely matching event occurrence rates, laboratory test results, and temporal dynamics in real patient future data. It also accurately estimated future event probabilities, with observed-to-expected ratios consistently near 1.0 across diverse outcomes and time horizons. Our results reveal the untapped value of real-world data in electronic health records and introduce a scalable framework for in silico modeling of clinical care.
引用
@article{arxiv.2601.17310,
title = {High-Fidelity Longitudinal Patient Simulation Using Real-World Data},
author = {Yu Akagi and Tomohisa Seki and Hiromasa Ito and Toru Takiguchi and Kazuhiko Ohe and Yoshimasa Kawazoe},
journal= {arXiv preprint arXiv:2601.17310},
year = {2026}
}