Communication Dans Un Congrès Année : 2025

ACUITEE: A Comprehensive Tool for Visualization, Editing and Curating textual Annotations in Clinical Data

Résumé

Annotation and management of clinical data remains a critical but challenging task due to the complexity and diversity of medical records. Providing a tool to simplify, shorten, and improve the annotation work of clinicians is essential as it leads both to process optimization and better patients' characterization. We present ACUITEE (Annotation and Curation User Interface for Terms Extraction Engines), a web application that addresses these challenges. It offers a simple way to improve clinical data annotation workflows by integrating automatic analysis, manual processing, and real-time visualization of medical notes. Using advanced natural language processing (NLP) techniques for phenotypes extraction such as PhenoBERT and efficient string-matching algorithms, ACUITEE maps free-text medical notes to ontology terms and enables clinicians to validate or refine these annotations through a user-friendly interface. The system supports fully automated, semi-automated and manual annotation modes, providing flexibility for different use cases. A key feature of ACUITEE is its interactive annotation interface, which enables clinicians to validate, edit, and curate ontology terms with precision, thereby speeding up the annotation process while maintaining high accuracy. This paper outlines ACUITEE's architecture, features, and applications and demonstrates its potential to facilitate clinical data annotation through increased efficiency, adaptability, and user engagement.

Fichier principal
Vignette du fichier
baddour2025icbes.pdf (1.15 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
licence

Dates et versions

hal-04974765 , version 1 (03-03-2025)

Licence

Identifiants

  • HAL Id : hal-04974765 , version 1

Citer

Moussa Baddour, Olivier Dameron, Marie de Tayrac, Stéphane Paquelet, Paul Rollier, et al.. ACUITEE: A Comprehensive Tool for Visualization, Editing and Curating textual Annotations in Clinical Data. EECSS 2025 - 11th World Congress on Electrical Engineering and Computer Systems and Sciences, Aug 2025, Paris, France. pp.1-9. ⟨hal-04974765⟩
870 Consultations
317 Téléchargements

Partager

  • More