ICD-Deepresearch：基础模型与深度研究结合的临床代码预测
原标题：Foundation Agents Meet Agentic Deep Research: Evidence-Grounded Clinical Code Forecasting
AI 摘要
arXiv 论文提出 ICD-Deepresearch，一种结合结构化 EHR 基础模型（SparseEHR）与语言模型（GPT-5）的深度研究流程，用于预测下一次就诊的 ICD 诊断代码。该方法通过候选生成、研究扩展和最终选择，在 MIMIC-III 和 MIMIC-IV 数据集上取得了优于基线（如 GPT-5 网页搜索和 Medical Deep Research）的性能，且医生评估其检索文档的有用性更高。
正文节选
Foundation Agents Meet Agentic Deep Research: Evidence-Grounded Clinical Code Forecasting Abstract Next-encounter ICD forecasting predicts which standardized diagnosis codes will be documented at a future visit from the longitudinal record available beforehand. The task is prospective and multi-label: the target note does not yet exist, and several codes may be correct. Structured EHR foundation models capture recurrence and temporal progression, whereas language foundation models generate flexi