Cover Image

Comparative analysis of machine learning algorithms for computer-assisted reporting based on fully automated cross-lingual RadLex mappings

Computer-assisted reporting (CAR) tools were suggested to improve radiology report quality by context-sensitively recommending key imaging biomarkers. However, studies evaluating machine learning (ML) algorithms on cross-lingual ontological (RadLex) mappings for developing embedded CAR algorithms ar...

Full description

Saved in:
Bibliographic Details
Main Authors: Maros, Máté E. (Author) , Cho, Chang Gyu (Author) , Junge, Andreas Georg (Author) , Kämpgen, Benedikt (Author) , Saase, Victor (Author) , Siegel, Fabian (Author) , Trinkmann, Frederik (Author) , Ganslandt, Thomas (Author) , Groden, Christoph (Author) , Wenz, Holger (Author)
Format: Article (Journal)
Language:English
Published: 09 March 2021
In: Scientific reports
Year: 2021, Volume: 11, Pages: 1-18
ISSN:2045-2322
DOI:10.1038/s41598-021-85016-9
Online Access:Verlag, kostenfrei, Volltext: https://doi.org/10.1038/s41598-021-85016-9
Verlag, kostenfrei, Volltext: https://www.nature.com/articles/s41598-021-85016-9
Get full text
Author Notes:Máté E. Maros, Chang Gyu Cho, Andreas G. Junge, Benedikt Kämpgen, Victor Saase, Fabian Siegel, Frederik Trinkmann, Thomas Ganslandt, Christoph Groden and Holger Wenz
Description
Summary:Computer-assisted reporting (CAR) tools were suggested to improve radiology report quality by context-sensitively recommending key imaging biomarkers. However, studies evaluating machine learning (ML) algorithms on cross-lingual ontological (RadLex) mappings for developing embedded CAR algorithms are lacking. Therefore, we compared ML algorithms developed on human expert-annotated features against those developed on fully automated cross-lingual (German to English) RadLex mappings using 206 CT reports of suspected stroke. Target label was whether the Alberta Stroke Programme Early CT Score (ASPECTS) should have been provided (yes/no:154/52). We focused on probabilistic outputs of ML-algorithms including tree-based methods, elastic net, support vector machines (SVMs) and fastText (linear classifier), which were evaluated in the same 5 × fivefold nested cross-validation framework. This allowed for model stacking and classifier rankings. Performance was evaluated using calibration metrics (AUC, brier score, log loss) and -plots. Contextual ML-based assistance recommending ASPECTS was feasible. SVMs showed the highest accuracies both on human-extracted- (87%) and RadLex features (findings:82.5%; impressions:85.4%). FastText achieved the highest accuracy (89.3%) and AUC (92%) on impressions. Boosted trees fitted on findings had the best calibration profile. Our approach provides guidance for choosing ML classifiers for CAR tools in fully automated and language-agnostic fashion using bag-of-RadLex terms on limited expert-labelled training data.
Item Description:Gesehen am 12.08.2021
Physical Description:Online Resource
ISSN:2045-2322
DOI:10.1038/s41598-021-85016-9

Similar Items