Deep Learning for Automated Classification of Hip Hardware on Radiographs
Journal of Imaging Informatics in Medicine · 12.09.2024 · 38 (2) · 988-996
DOI: 10.1007/s10278-024-01263-y ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Imaging Informatics in Medicine, online erschienen am 12.09.2024. Verfasst von Yuntong Ma und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, San Francisco VA Medical Center, 4150 Clement St, San Francisco, CA, 94121, USA. Hauptthemen laut MEDLINE: Deep Learning, Radiography, Hip Prosthesis, Hip Joint. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.
Abstract (Originalsprache)
PURPOSE: To develop a deep learning model for automated classification of orthopedic hardware on pelvic and hip radiographs, which can be clinically implemented to decrease radiologist workload and improve consistency among radiology reports. MATERIALS AND METHODS: Pelvic and hip radiographs from 4279 studies in 1073 patients were retrospectively obtained and reviewed by musculoskeletal radiologists. Two convolutional neural networks, EfficientNet-B4 and NFNet-F3, were trained to perform the image classification task into the following most represented categories: no hardware, total hip arthroplasty (THA), hemiarthroplasty, intramedullary nail, femoral neck cannulated screws, dynamic hip screw, lateral blade/plate, THA with additional femoral fixation, and post-infectious hip. Model performance was assessed on an independent test set of 851 studies from 262 patients and compared to individual performance of five subspecialty-trained radiologists using leave-one-out analysis against an aggregate gold standard label. RESULTS: For multiclass classification, the area under the receiver operating characteristic curve (AUC) for NFNet-F3 was 0.99 or greater for all classes, and EfficientNet-B4 0.99 or greater for all classes except post-infectious hip, with an AUC of 0.97. When compared with human observers, models achieved an accuracy of 97%, which is non-inferior to four out of five radiologists and outperformed one radiologist. Cohen's kappa coefficient for both models ranged from 0.96 to 0.97, indicating excellent inter-reader agreement. CONCLUSION: A deep learning model can be used to classify a range of orthopedic hip hardware with high accuracy and comparable performance to subspecialty-trained radiologists.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Radiology, San Francisco VA Medical Center, 4150 Clement St, San Francisco, CA, 94121, USA.
- Department of Radiology, Stanford Medicine. 300 Pasteur Dr, Palo Alto, CA, 94304, USA.
- Menlo-Atherton High School, 555 Middlefield Road Atherton, Atherton, CA, 94027, USA.
- Department of Radiology, Stanford Medicine. 300 Pasteur Dr, Palo Alto, CA, 94304, USA. cxfang@stanford.edu.
Zitieren
APA
Ma, Y., Bauer, J. L., Yoon, A. H., Beaulieu, C. F., Yoon, L., Do, B. H., & Fang, C. X. (2024). Deep Learning for Automated Classification of Hip Hardware on Radiographs. Journal of Imaging Informatics in Medicine, 38(2), 988-996. https://doi.org/10.1007/s10278-024-01263-y
Vancouver
Ma Y, Bauer JL, Yoon AH, Beaulieu CF, Yoon L, Do BH, et al. Deep Learning for Automated Classification of Hip Hardware on Radiographs. Journal of Imaging Informatics in Medicine. 2024;38(2):988-996. doi:10.1007/s10278-024-01263-y
BibTeX
@article{Ma2024,
author = {Ma, Yuntong and Bauer, Justin L and Yoon, Acacia H and Beaulieu, Christopher F and Yoon, Luke and Do, Bao H and Fang, Charles X},
title = {Deep Learning for Automated Classification of Hip Hardware on Radiographs},
journal = {Journal of Imaging Informatics in Medicine},
year = {2024},
volume = {38},
number = {2},
pages = {988--996},
doi = {10.1007/s10278-024-01263-y}
}
EUVIMED