HSMP-BERT, using a section-matching pretraining objective and hierarchical prompts, extracts organ-finding labels from Hebrew radiology reports with mean F1 0.83 and 5.1x faster inference.
Performance of large language models in numerical versus semantic medical knowledge: Cross-sectional benchmarking study on evidence-based questions and answers,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Hierarchical Section Matching Prediction (HSMP) BERT for Fine-Grained Extraction of Structured Data from Hebrew Free-Text Radiology Reports in Crohn's Disease
HSMP-BERT, using a section-matching pretraining objective and hierarchical prompts, extracts organ-finding labels from Hebrew radiology reports with mean F1 0.83 and 5.1x faster inference.