TY  - JOUR
AU  - Dada, Amin
AU  - Ufer, Tim Leon
AU  - Kim, Moon
AU  - Hasin, Max
AU  - Spieker, Nicola
AU  - Forsting, Michael
AU  - Nensa, Felix
AU  - Egger, Jan
AU  - Kleesiek, Jens
TI  - Information extraction from weakly structured radiological reports with natural language queries.
JO  - European radiology
VL  - 34
IS  - 1
SN  - 0938-7994
CY  - Heidelberg
PB  - Springer
M1  - DKFZ-2023-01528
SP  - 330-337
PY  - 2024
N1  - 2024 Jan;34(1):330-337
AB  - Provide physicians and researchers an efficient way to extract information from weakly structured radiology reports with natural language processing (NLP) machine learning models.We evaluate seven different German bidirectional encoder representations from transformers (BERT) models on a dataset of 857,783 unlabeled radiology reports and an annotated reading comprehension dataset in the format of SQuAD 2.0 based on 1223 additional reports.Continued pre-training of a BERT model on the radiology dataset and a medical online encyclopedia resulted in the most accurate model with an F1-score of 83.97
KW  - Information extraction (Other)
KW  - Machine learning (Other)
KW  - Natural language processing (Other)
LB  - PUB:(DE-HGF)16
C6  - pmid:37505252
DO  - DOI:10.1007/s00330-023-09977-3
UR  - https://inrepo02.dkfz.de/record/277860
ER  -