Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MedTri: A Platform for Structured Medical Report Normalization to Enhance Vision-Language Pretraining

Yuetan Chu, Xinhua Ma, Xinran Jin, Gongning Luo, Xin Gao

TL;DR

This study presents MedTri, a deployable normalization framework for medical vision-language pretraining that converts free-text reports into a unified [Anatomical Entity: Radiologic Description + Diagnosis Category] triplet, and illustrates how this normalization can easily support modular text-level augmentation strategies, including knowledge enrichment and anatomy-grounded counterfactual supervision.

Abstract

Medical vision-language pretraining increasingly relies on medical reports as large-scale supervisory signals; however, raw reports often exhibit substantial stylistic heterogeneity, variable length, and a considerable amount of image-irrelevant content. Although text normalization is frequently adopted as a preprocessing step in prior work, its design principles and empirical impact on vision-language pretraining remain insufficiently and systematically examined. In this study, we present MedTri, a deployable normalization framework for medical vision-language pretraining that converts free-text reports into a unified [Anatomical Entity: Radiologic Description + Diagnosis Category] triplet. This structured, anatomy-grounded normalization preserves essential morphological and spatial information while removing stylistic noise and image-irrelevant content, providing consistent and image-grounded textual supervision at scale. Across multiple datasets spanning both X-ray and computed tomography (CT) modalities, we demonstrate that structured, anatomy-grounded text normalization is an important factor in medical vision-language pretraining quality, yielding consistent improvements over raw reports and existing normalization baselines. In addition, we illustrate how this normalization can easily support modular text-level augmentation strategies, including knowledge enrichment and anatomy-grounded counterfactual supervision, which provide complementary gains in robustness and generalization without altering the core normalization process. Together, our results position structured text normalization as a critical and generalizable preprocessing component for medical vision-language learning, while MedTri provides this normalization platform. Code and data will be released at https://github.com/Arturia-Pendragon-Iris/MedTri.

MedTri: A Platform for Structured Medical Report Normalization to Enhance Vision-Language Pretraining

TL;DR

This study presents MedTri, a deployable normalization framework for medical vision-language pretraining that converts free-text reports into a unified [Anatomical Entity: Radiologic Description + Diagnosis Category] triplet, and illustrates how this normalization can easily support modular text-level augmentation strategies, including knowledge enrichment and anatomy-grounded counterfactual supervision.

Abstract

Medical vision-language pretraining increasingly relies on medical reports as large-scale supervisory signals; however, raw reports often exhibit substantial stylistic heterogeneity, variable length, and a considerable amount of image-irrelevant content. Although text normalization is frequently adopted as a preprocessing step in prior work, its design principles and empirical impact on vision-language pretraining remain insufficiently and systematically examined. In this study, we present MedTri, a deployable normalization framework for medical vision-language pretraining that converts free-text reports into a unified [Anatomical Entity: Radiologic Description + Diagnosis Category] triplet. This structured, anatomy-grounded normalization preserves essential morphological and spatial information while removing stylistic noise and image-irrelevant content, providing consistent and image-grounded textual supervision at scale. Across multiple datasets spanning both X-ray and computed tomography (CT) modalities, we demonstrate that structured, anatomy-grounded text normalization is an important factor in medical vision-language pretraining quality, yielding consistent improvements over raw reports and existing normalization baselines. In addition, we illustrate how this normalization can easily support modular text-level augmentation strategies, including knowledge enrichment and anatomy-grounded counterfactual supervision, which provide complementary gains in robustness and generalization without altering the core normalization process. Together, our results position structured text normalization as a critical and generalizable preprocessing component for medical vision-language learning, while MedTri provides this normalization platform. Code and data will be released at https://github.com/Arturia-Pendragon-Iris/MedTri.
Paper Structure (13 sections, 1 equation, 2 figures, 5 tables)

This paper contains 13 sections, 1 equation, 2 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Method
  3. Structured Triplet
  4. Local Model Development
  5. Optional Text-Level Augmentation on MedTri
  6. Medical Knowledge Expansion (MedTri-K)
  7. Anatomy-Grounded Counterfactuals (MedTri-C)
  8. Experiments and Results
  9. Evaluation of MedTri for Normalization
  10. MedTri for Improving Downstream Tasks
  11. Experiment Settings and Datasets
  12. Downstream Task Performances
  13. Discussion and Conclusion

Figures (2)

  • Figure 1: Illustration of normalization differences across raw reports, RadGraph, and MedTri. The raw clinical report contains irrelevant content and mixed structures. RadGraph extracts only diagnostic entities with limited imaging descriptions. MedTri produces anatomically anchored, image-grounded triplets that preserve morphological and spatial detail while removing stylistic and clinically irrelevant text.
  • Figure 2: Illustration of MedTri’s optional augmentation modules. MedTri-K enriches each triplet with clinically validated radiologic signatures to improve visual interpretability. MedTri-C generates anatomically inconsistent counterfactuals through controlled perturbations to strengthen fine-grained visual–semantic discrimination.