Using BERT models for breast cancer diagnosis from Turkish radiology reports

Uskaner Hepsağ P., Özel S. A., Dalcı K., Yazıcı A.

Language Resources and Evaluation, cilt.58, sa.3, ss.981-1012, 2024 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 58 Sayı: 3
  • Basım Tarihi: 2024
  • Doi Numarası: 10.1007/s10579-023-09669-w
  • Dergi Adı: Language Resources and Evaluation
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, FRANCIS, Aerospace Database, Applied Science & Technology Source, Communication Abstracts, Computer & Applied Sciences, EBSCO Education Source, Educational research abstracts (ERA), Humanities Abstracts, INSPEC, Linguistic Bibliography, Linguistics & Language Behavior Abstracts, Metadex, MLA - Modern Language Association Database, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.981-1012
  • Anahtar Kelimeler: Breast cancer, Contextualized word embeddings, Machine learning, Radiology reports, Turkish dataset
  • Çukurova Üniversitesi Adresli: Evet

Özet

Diagnostic radiology is concerned with obtaining images of the internal organs using radiological imaging procedures. These images are then interpreted by a diagnostic radiologist, who produces a textual report that assists in the diagnosis of illness or injury. Early detection of certain illnesses, particularly cancer, is critical, and the reports produced by diagnostic radiologists play a key role in this process. To develop models for the early detection of cancer, text classification techniques can be applied to radiological reports. However, this process requires access to a dataset of radiology reports, which is not widely available. Currently, radiology report datasets exist for high-resource languages such as English and Dutch, but not for low-resource languages such as Turkish. This article describes the collection of a mammography report dataset for Turkish, consisting of 62 reports from real patients that were manually labeled by an expert for diagnosing breast cancer. Basic machine learning models were applied to this dataset using pre-trained BERT, DistilBERT, and an ensemble learning hard voting approach. The results showed that BERT on Turkish achieved the best performance, with a 91% F1-score. Hard Voting, which combined the results of BERTTurkish, BERTClinical, and BERTMultilingual, achieved the highest F1-score of 93%. The results show that BERT and Hard Voting outperform the other machine learning models for breast cancer diagnosis from Turkish radiology reports.