CARES: A Corpus for classification of Spanish Radiological reports. (March 2023)

• Record Type: Journal Article
• Authors: Chizhikova, Mariia; López-Úbeda, Pilar; Collado-Montañez, Jaime; Martín-Noguerol, Teodoro; Díaz-Galiano, Manuel C.; Luna, Antonio; Ureña-López, L. Alfonso; Martín-Valdivia, M. Teresa
• Is Part Of: Computers in biology and medicine. Volume 154(2023)
• Publication Date: 2023-03
• Subjects: Biomedical natural language processing -- Radiology report -- Text classification -- Medical corpus -- ICD-10 -- Transformer language model -- Spanish medical resource; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2023.106581
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: This paper presents a new corpus of radiology medical reports written in Spanish and labeled with ICD-10. CARES (Corpus of Anonymised Radiological Evidences in Spanish) is a high-quality corpus manually labeled and reviewed by radiologists that is freely available for the research community on HuggingFace. These types of resources are essential for developing automatic text classification tools as they are necessary for training and tuning computational systems. However, in the medical domain these are very difficult to obtain for different reasons including privacy and data protection issues or the involvement of medical specialists in the generation of these resources. We present a corpus labeled and reviewed by radiologists in their daily practice that is available for research purposes. In addition, after describing the corpus and explaining how it has been generated, a first experimental approach is carried out using several machine learning algorithms based on transformer language models such as BioBERT and RoBERTa to test the validity of this linguistic resource. The best performing classifier achieved 0.8676 micro and 0.8328 macro f1-score and these results encourage us to continue working in this research line. Highlights: A freely-available corpus of radiological reports with ICD-10 codes is presented Reports labeled with Z03.9 sub-code enable systems to classify normality cases Report classification into ICD-10 chapters is tackled with 0.83 macro F1-score