Saved in:
Bibliographic Details
Main Authors: Yu, Yi, Martin, Parker, Bu, Zhenyu, Liu, Yixuan, Zheng, Yi-Yu, Simonetti, Orlando, Han, Yuchi, Xue, Yuan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.08045
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918489907789824
author Yu, Yi
Martin, Parker
Bu, Zhenyu
Liu, Yixuan
Zheng, Yi-Yu
Simonetti, Orlando
Han, Yuchi
Xue, Yuan
author_facet Yu, Yi
Martin, Parker
Bu, Zhenyu
Liu, Yixuan
Zheng, Yi-Yu
Simonetti, Orlando
Han, Yuchi
Xue, Yuan
contents Converting free-text cardiac magnetic resonance (CMR) reports into auditable structured data remains a bottleneck for cohort assembly, longitudinal curation, and clinical decision support. We present CMR-EXTR, a lightweight framework that converts free-text CMR reports into structured data and assigns per-field confidence for quality control. A teacher-student distillation pipeline enables fully offline inference while limiting manual annotation. Uncertainty integrates three complementary principles -- distribution plausibility, sampling stability, and cross-field consistency -- to triage human review. Experiments show that CMR-EXTR achieves 99.65% variable-level accuracy, demonstrating both reliable extraction and informative confidence scores. To our knowledge, this is the first CMR-specific extraction system with integrated confidence estimation. The code is available at https://github.com/yuyi1005/CMR-EXTR.
format Preprint
id arxiv_https___arxiv_org_abs_2605_08045
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Uncertainty-Aware Structured Data Extraction from Full CMR Reports via Distilled LLMs
Yu, Yi
Martin, Parker
Bu, Zhenyu
Liu, Yixuan
Zheng, Yi-Yu
Simonetti, Orlando
Han, Yuchi
Xue, Yuan
Computation and Language
Converting free-text cardiac magnetic resonance (CMR) reports into auditable structured data remains a bottleneck for cohort assembly, longitudinal curation, and clinical decision support. We present CMR-EXTR, a lightweight framework that converts free-text CMR reports into structured data and assigns per-field confidence for quality control. A teacher-student distillation pipeline enables fully offline inference while limiting manual annotation. Uncertainty integrates three complementary principles -- distribution plausibility, sampling stability, and cross-field consistency -- to triage human review. Experiments show that CMR-EXTR achieves 99.65% variable-level accuracy, demonstrating both reliable extraction and informative confidence scores. To our knowledge, this is the first CMR-specific extraction system with integrated confidence estimation. The code is available at https://github.com/yuyi1005/CMR-EXTR.
title Uncertainty-Aware Structured Data Extraction from Full CMR Reports via Distilled LLMs
topic Computation and Language
url https://arxiv.org/abs/2605.08045