2019/08/13 by Surabhi Datta, Datta, Surabhi, Yuqi Si +9 · 3 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · #Biomedical Text Mining and Ontologies #Computation and Language (cs.CL) #FOS: Computer and information sciences #Natural Language Processing Techniques #Topic Modeling
paper · pdf · doi:10.48550/arxiv.1908.04485
openalex publication_date 2019/08/13 · openalex created_date 2022/07/28 · openalex updated_date 2026/07/28
We define a representation framework for extracting spatial information from\nradiology reports (Rad-SpRL). We annotated a total of 2000 chest X-ray reports\nwith 4 spatial roles corresponding to the common radiology entities. Our focus\nis on extracting detailed information of a radiologist's interpretation\ncontaining a radiographic finding, its anatomical location, corresponding\nprobable diagnoses, as well as associated hedging terms. For this, we propose a\ndeep learning-based natural language processing (NLP) method involving both\nword and character-level encodings. Specifically, we utilize a bidirectional\nlong short-term memory (Bi-LSTM) conditional random field (CRF) model for\nextracting the spatial roles. The model achieved average F1 measures of 90.28\nand 94.61 for extracting the Trajector and Landmark roles respectively whereas\nthe performance was moderate for Diagnosis and Hedge roles with average F1 of\n71.47 and 73.27 respectively. The corpus will soon be made available upon\nrequest.\n