BERT-based Transfer Learning in Sentence-level Anatomic Classification of Free-Text Radiology Reports

Daiki Nishigaki, Yuki Suzuki, Tomohiro Wataya, Kosuke Kita, Kazuki Yamagata, Junya Sato, Shoji Kido, Noriyuki Tomiyama

doi:10.1148/ryai.220097

To assess whether transfer learning with a bidirectional encoder representations from transformers (BERT) model, pretrained on a clinical corpus, can perform sentence-level anatomic classification of free-text radiology reports, even for anatomic classes with few positive examples.This retrospective study included radiology reports of patients who underwent whole-body PET/CT imaging from December 2005 to December 2020. Each sentence in these reports (6272 sentences) was labeled by two annotators according to body part ("brain," "headneck," "chest," "abdomen," "limbs," "spine," or "others"). The BERT-based transfer learning approach was compared with two baseline machine learning approaches: bidirectional long short-term memory (BiLSTM) and the count-based method. Area under the precision-recall curve (AUPRC) and area under the receiver operating characteristic curve (AUC) were computed for each approach, and AUCs were compared using the DeLong test.The BERT-based approach achieved a macro-averaged AUPRC of 0.88 for classification, outperforming the baselines. AUC results for BERT were significantly higher than those of BiLSTM for all classes and those of the count-based method for the "brain," "chest," "abdomen," and "others" classes (The BERT-based transfer learning approach outperformed the BiLSTM and count-based approaches in sentence-level anatomic classification of free-text radiology reports, even for anatomic classes with few labeled training data.

BERT-based Transfer Learning in Sentence-level Anatomic Classification of Free-Text Radiology Reports

Description

Journal

Citations (1)*help

Details 詳細情報について

Export

Report a problem