EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Improving Radiology Report Generation Quality and Diversity through Reinforcement Learning and Text Augmentation

Daniel Parres iD, Alberto Albiol iD, Roberto Paredes iD

Bioengineering · 03.04.2024 · 11 (4) · 351

DOI: 10.3390/bioengineering11040351 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bioengineering, online erschienen am 03.04.2024. Verfasst von Daniel Parres und 2 weiteren Autor:innen. Beteiligte Einrichtung: Campus de Vera, Universitat Politècnica València, Camí de Vera s/n, 46022 Valencia, Spain. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Generalitat Valenciana.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Deep learning is revolutionizing radiology report generation (RRG) with the adoption of vision encoder–decoder (VED) frameworks, which transform radiographs into detailed medical reports. Traditional methods, however, often generate reports of limited diversity and struggle with generalization. Our research introduces reinforcement learning and text augmentation to tackle these issues, significantly improving report quality and variability. By employing RadGraph as a reward metric and innovating in text augmentation, we surpass existing benchmarks like BLEU4, ROUGE-L, F1CheXbert, and RadGraph, setting new standards for report accuracy and diversity on MIMIC-CXR and Open-i datasets. Our VED model achieves F1-scores of 66.2 for CheXbert and 37.8 for RadGraph on the MIMIC-CXR dataset, and 54.7 and 45.6, respectively, on Open-i. These outcomes represent a significant breakthrough in the RRG field. The findings and implementation of the proposed approach, aimed at enhancing diagnostic precision and radiological interpretations in clinical settings, are publicly available on GitHub to encourage further advancements in the field.

Schlagwörter

Stichwörter der Autor:innen

chest X-raysdeep learningmachine learningmedical imageradiology report generationreinforcement learningtext augmentationtext generationvision transformer

Autor:innen und Institutionen

  • Campus de Vera, Universitat Politècnica València, Camí de Vera s/n, 46022 Valencia, Spain.
  • Valencian Graduate School and Research Network of Artificial Intelligence, Camí de Vera s/n, 46022 Valencia, Spain.

ORCID-Profile: Daniel Parres, Alberto Albiol, Roberto Paredes

Förderung

  • Generalitat Valenciana

Zitieren

APA

Parres, D., Albiol, A., & Paredes, R. (2024). Improving Radiology Report Generation Quality and Diversity through Reinforcement Learning and Text Augmentation. Bioengineering, 11(4), 351. https://doi.org/10.3390/bioengineering11040351

Vancouver

Parres D, Albiol A, Paredes R. Improving Radiology Report Generation Quality and Diversity through Reinforcement Learning and Text Augmentation. Bioengineering. 2024;11(4):351. doi:10.3390/bioengineering11040351

BibTeX
@article{Parres2024,
  author = {Parres, Daniel and Albiol, Alberto and Paredes, Roberto},
  title = {Improving Radiology Report Generation Quality and Diversity through Reinforcement Learning and Text Augmentation},
  journal = {Bioengineering},
  year = {2024},
  volume = {11},
  number = {4},
  pages = {351},
  doi = {10.3390/bioengineering11040351}
}
Kostenlos starten →