Weijuan Chen1, Xi Liu1, Kunhua Li1, Yin Luo1, Shanwei Bai1, Jiangfen Wu2, Weidao Chen2, Mengxing Dong2, Dajing Guo3. 1. Department of Radiology, The Second Affiliated Hospital of Chongqing Medical University, No. 74 Linjiang Rd, Yuzhong District, Chongqing, 400010, China. 2. Department of Applied Clinical Medicine, Infervision, Beijing, China. 3. Department of Radiology, The Second Affiliated Hospital of Chongqing Medical University, No. 74 Linjiang Rd, Yuzhong District, Chongqing, 400010, China. guodaj@hospital.cqmu.edu.cn.
Abstract
OBJECTIVES: To develop a deep-learning (DL) model for identifying fresh VCFs from digital radiography (DR), with magnetic resonance imaging (MRI) as the reference standard. METHODS: Patients with lumbar VCFs were retrospectively enrolled from January 2011 to May 2020. All patients underwent DR and MRI scanning. VCFs were categorized as fresh or old according to MRI results, and the VCF grade and type were assessed. The raw DR data were sent to InferScholar Center for annotation. A DL-based prediction model was built, and its diagnostic performance was evaluated. The DeLong test was applied to assess differences in ROC curves between different models. RESULTS: A total of 1877 VCFs in 1099 patients were included in our study and randomly divided into development (n = 824 patients) and test (n = 275 patients) datasets. The ensemble model identified fresh and old VCFs, reaching an AUC of 0.80 (95% confidence interval [CI], 0.77-0.83), an accuracy of 74% (95% CI, 72-77%), a sensitivity of 80% (95% CI, 77-83%), and a specificity of 68% (95% CI, 63-72%). Lateral (AUC, 0.83) views exhibited better performance than anteroposterior views (AUC, 0.77), and the best performance among respective subgroupings was obtained for grade 3 (AUC, 0.89) and crush-type (AUC, 0.87) subgroups. CONCLUSION: The proposed DL model achieved adequate performance in identifying fresh VCFs from DR. KEY POINTS: • The ensemble deep-learning model identified fresh VCFs from DR, reaching an AUC of 0.80, an accuracy of 74%, a sensitivity of 80%, and a specificity of 68% with the reference standard of MRI. • The lateral views (AUC, 0.83) exhibited better performance than anteroposterior views (AUC, 0.77). • The grade 3 (AUC, 0.89) and crush-type (AUC, 0.87) subgroups showed the best performance among their respective subgroupings.
OBJECTIVES: To develop a deep-learning (DL) model for identifying fresh VCFs from digital radiography (DR), with magnetic resonance imaging (MRI) as the reference standard. METHODS: Patients with lumbar VCFs were retrospectively enrolled from January 2011 to May 2020. All patients underwent DR and MRI scanning. VCFs were categorized as fresh or old according to MRI results, and the VCF grade and type were assessed. The raw DR data were sent to InferScholar Center for annotation. A DL-based prediction model was built, and its diagnostic performance was evaluated. The DeLong test was applied to assess differences in ROC curves between different models. RESULTS: A total of 1877 VCFs in 1099 patients were included in our study and randomly divided into development (n = 824 patients) and test (n = 275 patients) datasets. The ensemble model identified fresh and old VCFs, reaching an AUC of 0.80 (95% confidence interval [CI], 0.77-0.83), an accuracy of 74% (95% CI, 72-77%), a sensitivity of 80% (95% CI, 77-83%), and a specificity of 68% (95% CI, 63-72%). Lateral (AUC, 0.83) views exhibited better performance than anteroposterior views (AUC, 0.77), and the best performance among respective subgroupings was obtained for grade 3 (AUC, 0.89) and crush-type (AUC, 0.87) subgroups. CONCLUSION: The proposed DL model achieved adequate performance in identifying fresh VCFs from DR. KEY POINTS: • The ensemble deep-learning model identified fresh VCFs from DR, reaching an AUC of 0.80, an accuracy of 74%, a sensitivity of 80%, and a specificity of 68% with the reference standard of MRI. • The lateral views (AUC, 0.83) exhibited better performance than anteroposterior views (AUC, 0.77). • The grade 3 (AUC, 0.89) and crush-type (AUC, 0.87) subgroups showed the best performance among their respective subgroupings.
Authors: Moritz Kaup; Julian L Wichmann; Jan-Erik Scholtz; Martin Beeres; Wolfgang Kromen; Moritz H Albrecht; Thomas Lehnert; Marie Boettcher; Thomas J Vogl; Ralf W Bauer Journal: Radiology Date: 2016-02-29 Impact factor: 11.105
Authors: Torsten Diekhoff; Nils Engelhard; Michael Fuchs; Matthias Pumberger; Michael Putzier; Jürgen Mews; Marcus Makowski; Bernd Hamm; Kay-Geert A Hermann Journal: Eur Radiol Date: 2018-06-15 Impact factor: 5.315
Authors: Bernhard Petritsch; Aleksander Kosmala; Andreas M Weng; Bernhard Krauss; Anke Heidemeier; Richard Wagner; Timo M Heintel; Tobias Gassenmaier; Thorsten A Bley Journal: Radiology Date: 2017-02-27 Impact factor: 11.105
Authors: William Clark; Paul Bird; Peter Gonski; Terrence H Diamond; Peter Smerdely; H Patrick McNeil; Glen Schlaphoff; Carl Bryant; Elizabeth Barnes; Val Gebski Journal: Lancet Date: 2016-08-17 Impact factor: 79.321
Authors: Claudia Frellesen; Mehrnoush Azadegan; Simon S Martin; Katharina Otani; Tommaso DʼAngelo; Christian Booz; Katrin Eichler; Bita Panahi; Moritz Kaup; Ralf W Bauer; Thomas J Vogl; Julian L Wichmann Journal: Invest Radiol Date: 2018-07 Impact factor: 6.016
Authors: Maximilian T Löffler; Alina Jacob; Alexander Valentinitsch; Anna Rienmüller; Claus Zimmer; Yu-Mi Ryang; Thomas Baum; Jan S Kirschke Journal: Eur Radiol Date: 2019-02-21 Impact factor: 5.315
Authors: Douglas P Beall; M R Chambers; Sam Thomas; John Amburgy; James R Webb; Bradly S Goodman; Devin K Datta; Richard W Easton; Douglas Linville; Sanjay Talati; John B Tillman Journal: Neurosurgery Date: 2019-01-01 Impact factor: 4.654
Authors: Xiao Chen; Qingshan Deng; Qiang Wang; Xinmiao Liu; Lei Chen; Jinjin Liu; Shuangquan Li; Meihao Wang; Guoquan Cao Journal: Front Public Health Date: 2022-04-26