HarDNet-based deep learning model for osteoporosis screening and bone mineral density inference from hand radiographs

Chan-Shien Ho; Tzuo-Yau Fan; Chang-Fu Kuo; Tzu-Yun Yen; Szu-Yi Chang; Yu-Cheng Pei; Yueh-Peng Chen

doi:10.1016/j.bone.2024.117317

HarDNet-based deep learning model for osteoporosis screening and bone mineral density inference from hand radiographs

Bone. 2024 Nov 3:117317. doi: 10.1016/j.bone.2024.117317. Online ahead of print.

Authors

Chan-Shien Ho¹, Tzuo-Yau Fan², Chang-Fu Kuo³, Tzu-Yun Yen⁴, Szu-Yi Chang⁵, Yu-Cheng Pei⁶, Yueh-Peng Chen⁷

Affiliations

¹ Department of Physical Medicine and Rehabilitation, Taoyuan Chang Gung Memorial Hospital, Taoyuan, Taiwan; Department of Physical Medicine and Rehabilitation, No. 123, Dinghu Rd., Guishan Dist., Taoyuan City 333008, Taiwan; Comprehensive Sports Medicine Center, Taoyuan Chang Gung Memorial Hospital, Taoyuan, Taiwan; Comprehensive Sports Medicine Center, No. 123, Dinghu Rd., Guishan Dist., Taoyuan City 333008, Taiwan; Master of Science Degree Program in Innovation for Smart Medicine, Chang Gung University, Taoyuan, Taiwan; College of Management, No.259, Wenhua 1st Rd., Guishan Dist., Taoyuan City 33302, Taiwan; College of Management, Chang Gung University, Taoyuan, Taiwan.
² Department of Research and Development, Chang Gung Medical Technology Co., Ltd., No. 11-5, Wenhua 2nd Road., Ltd., Guishan District., Taoyuan City 333, Taiwan.
³ Division of Rheumatology, Allergy and Immunology, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Division of Rheumatology, Allergy and Immunology, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan; Center for Artificial Intelligence in Medicine, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Center for Artificial Intelligence in Medicine, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan; School of Medicine, Chang Gung University, Taoyuan, Taiwan; School of Medicine, No.259, Wenhua 1st Rd., Guishan Dist., Taoyuan City 33302, Taiwan.
⁴ Department of Physical Medicine and Rehabilitation, Taoyuan Chang Gung Memorial Hospital, Taoyuan, Taiwan; Department of Physical Medicine and Rehabilitation, No. 123, Dinghu Rd., Guishan Dist., Taoyuan City 333008, Taiwan.
⁵ Center for Artificial Intelligence in Medicine, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Center for Artificial Intelligence in Medicine, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan.
⁶ Master of Science Degree Program in Innovation for Smart Medicine, Chang Gung University, Taoyuan, Taiwan; College of Management, No.259, Wenhua 1st Rd., Guishan Dist., Taoyuan City 33302, Taiwan; School of Medicine, Chang Gung University, Taoyuan, Taiwan; School of Medicine, No.259, Wenhua 1st Rd., Guishan Dist., Taoyuan City 33302, Taiwan; Department of Physical Medicine and Rehabilitation, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Department of Physical Medicine and Rehabilitation, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan; Center of Vascularized Tissue Allograft, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Center of Vascularized Tissue Allograft, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan. Electronic address: [email protected].
⁷ Master of Science Degree Program in Innovation for Smart Medicine, Chang Gung University, Taoyuan, Taiwan; College of Management, No.259, Wenhua 1st Rd., Guishan Dist., Taoyuan City 33302, Taiwan; Division of Rheumatology, Allergy and Immunology, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Division of Rheumatology, Allergy and Immunology, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan; Center for Artificial Intelligence in Medicine, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan; Center for Artificial Intelligence in Medicine, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan. Electronic address: [email protected].

PMID: 39500404
DOI: 10.1016/j.bone.2024.117317

Abstract

Purpose: Osteoporosis, affecting over 200 million individuals, often remains unrecognized and untreated, increasing the risk of fractures in older adults. Osteoporosis is typically diagnosed with bone mineral density (BMD) measured by dual-energy X-ray absorptiometry (DXA). This study aims to develop DeepDXA-Hand, a deep learning model using the efficient CNN-based deep learning architecture, for opportunistic osteoporosis screening from hand radiographs.

Methods: DeepDXA-Hand utilizes a CNN-based, HarDNet, approach to predict BMD non-invasively. A total of 10,351 hand radiographs and DXA pairs were used for model training and validation. The model's interpretability was enhanced using GradCAM for hotspot analysis to determine the model's attention areas.

Results: The predicted and ground truth BMD were significantly correlated with a correlation coefficient of 0.745. For binary classification of osteoporosis, DeepDXA-Hand demonstrated a sensitivity of 0.73, specificity of 0.83, and accuracy of 0.80, indicating its clinical potential. The model mainly focused on the carpal bones, such as the capitate, trapezoid, hamate, triquetrum, and the head of the second metacarpal bone, suggesting these areas provide radiological features for inferring BMD.

Conclusion: DeepDXA-Hand shows potential for the early detection of osteoporosis with high sensitivity and specificity. Further studies should explore its utility in predicting fracture risks.

Mini abstract: Osteoporosis affects millions and often goes undetected and untreated. DeepDXA-Hand, a HarDNet-based deep learning model, predicted bone mineral density with a correlation of 0.745 and classified osteoporosis with 0.80 accuracy. This model enhances early detection and has significant clinical potential as osteoporosis opportunistic screening tool.

Keywords: Bone mineral density (BMD); Diagnostic imaging; Hand radiographs; HarDNet model; Opportunistic screening; Osteoporosis.