Scholarlink

Extracorporeal Membrane Oxygenation (ECMO) is a method for supporting patients with severe cardiac or respiratory failure. However, pediatric ECMO patients are at a higher risk of severe neurological injury (NI). Understanding underlying causal mechanisms is critical for clinical decision-making. To this effect, we explore using Large Language Models (LLMs) for the construction of Causal Bayesian networks. While LLMs can reproduce causal relationships reflected in their training data, they may also generate spurious associations. We address this by refining the LLM-generated BN using data from 71 patients and domain constraints elicited from our experts. Our empirical evaluation shows that our method can construct causal diagrams by combining domain knowledge with empirical patterns.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

LLM-Guided Causal Bayesian Network Construction for Pediatric Patients on ECMO

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Saurabh Mathur,
  • Ranveer Singh,
  • Michael Skinner,
  • Ethan Sanford,
  • Neel Shah,
  • Phillip Reeder,
  • Lakshmi Raman,
  • Sriraam Natarajan

摘要

Extracorporeal Membrane Oxygenation (ECMO) is a method for supporting patients with severe cardiac or respiratory failure. However, pediatric ECMO patients are at a higher risk of severe neurological injury (NI). Understanding underlying causal mechanisms is critical for clinical decision-making. To this effect, we explore using Large Language Models (LLMs) for the construction of Causal Bayesian networks. While LLMs can reproduce causal relationships reflected in their training data, they may also generate spurious associations. We address this by refining the LLM-generated BN using data from 71 patients and domain constraints elicited from our experts. Our empirical evaluation shows that our method can construct causal diagrams by combining domain knowledge with empirical patterns.