Datasets

• MtSamples
  • Medical Transcriptions
  • Find the annotated versions by three different annotators under data/processed/mtsamples
• UMLS
  • MRSCONSO.RRF file
  • Put under data/raw
• ICD10 2017
  • icd10cm_order_2017.txt file
  • Put under data/raw
• big.txt
  • variety of ebooks which are merged into one document
  • Project Gutenberg's The Adventures of Sherlock Holmes, by Arthur Conan Doyle http://www.gutenberg.org/files/1661/1661-0.txt
  • The Project Gutenberg EBook of History of the United States by Charles A. Beard and Mary R. Beard http://www.gutenberg.org/cache/epub/16960/pg16960.txt
  • The Project Gutenberg EBook of War and Peace, by Leo Tolstoy http://www.gutenberg.org/files/2600/2600-0.txt
  • Put under data/raw

Word vectors

• Word2Vec
  • Put under data/raw/embeddings/
• BioBERT
  • BERT embeddings trained on PubMed dataset

Experiments

• Tune models: train/tune_models.py

  • Tunes dataset-model pairs using grid search
  • Saves results to Wandb

• Generate model performance reports: train/run_reports.py

  • Loads the best dataset-model pair and related config file of a trained model from wandb
  • Generates and saves results figures

Citation

@article{ozyegen2022word,
  title={Word-level text highlighting of medical texts for telehealth services},
  author={Ozyegen, Ozan and Kabe, Devika and Cevik, Mucahit},
  journal={Artificial Intelligence in Medicine},
  pages={102284},
  year={2022},
  publisher={Elsevier}
}