Papers/2608.17075
🧪 Test?View on arXiv

Foundation Agents Meet Agentic Deep Research: Evidence-Grounded Clinical Code Forecasting

Not specified in the provided content

RAGmulti-label forecastingEHRclinical coding
2608.17075
Builder Relevance
80%
2h ago

Abstract

The paper presents ICD-Deepresearch, a workflow for predicting future ICD codes using structured EHR models and language foundation models.

Reality Card

Core Claim

ICD-Deepresearch achieves improved patient-averaged precision/recall compared to standalone systems, demonstrating higher physician-rated usefulness.

Method / Result

Achieved patient-averaged precision/recall of 24.60/35.09% on MIMIC-III and 25.14/48.32% on MIMIC-IV.

Limitations

The reliance on fixed top-K budget and the specific datasets may limit generalizability and reproducibility.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers