Yijie Fang1,2, Wei Li1,2, Xiaojun Chen1,2, Keming Chen1,2, Han Kang3, Pengxin Yu3, Rongguo Zhang3, Jianwei Liao1,2, Guobin Hong4,5, Shaolin Li6,7. 1. Department of Radiology, The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, 519000, Guangdong Province, China. 2. Guangdong Provincial Key Laboratory of Biomedical Imaging, The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, 519000, Guangdong Province, China. 3. Institute of Advanced Research, Infervision, Beijing, 100025, China. 4. Department of Radiology, The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, 519000, Guangdong Province, China. honggb@mail.sysu.edu.cn. 5. Guangdong Provincial Key Laboratory of Biomedical Imaging, The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, 519000, Guangdong Province, China. honggb@mail.sysu.edu.cn. 6. Department of Radiology, The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, 519000, Guangdong Province, China. Lishlin5@mail.sysu.edu.cn. 7. Guangdong Provincial Key Laboratory of Biomedical Imaging, The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, 519000, Guangdong Province, China. Lishlin5@mail.sysu.edu.cn.
Abstract
OBJECTIVE: To explore the application of deep learning in patients with primary osteoporosis, and to develop a fully automatic method based on deep convolutional neural network (DCNN) for vertebral body segmentation and bone mineral density (BMD) calculation in CT images. MATERIALS AND METHODS: A total of 1449 patients were used for experiments and analysis in this retrospective study, who underwent spinal or abdominal CT scans for other indications between March 2018 and May 2020. All data was gathered from three different CT vendors. Among them, 586 cases were used for training, and other 863 cases were used for testing. A fully convolutional neural network, called U-Net, was employed for automated vertebral body segmentation. The manually sketched region of vertebral body was used as the ground truth for comparison. A convolutional neural network, called DenseNet-121, was applied for BMD calculation. The values post-processed by quantitative computed tomography (QCT) were identified as the standards for analysis. RESULTS: Based on the diversity of CT vendors, all testing cases were split into three testing cohorts: Test set 1 (n = 463), test set 2 (n = 200), and test set 3 (n = 200). Automated segmentation correlated well with manual segmentation regarding four lumbar vertebral bodies (L1-L4): the minimum average dice coefficients for three testing sets were 0.823, 0.786, and 0.782, respectively. For testing sets from different vendors, the average BMDs calculated by automated regression showed high correlation (r > 0.98) and agreement with those derived from QCT. CONCLUSIONS: A deep learning-based method could achieve fully automatic identification of osteoporosis, osteopenia, and normal bone mineral density in CT images. KEY POINTS: • Deep learning can perform accurate fully automated segmentation of lumbar vertebral body in CT images. • The average BMDs obtained by deep learning highly correlates with ones derived from QCT. • The deep learning-based method could be helpful for clinicians in opportunistic osteoporosis screening in spinal or abdominal CT scans.
OBJECTIVE: To explore the application of deep learning in patients with primary osteoporosis, and to develop a fully automatic method based on deep convolutional neural network (DCNN) for vertebral body segmentation and bone mineral density (BMD) calculation in CT images. MATERIALS AND METHODS: A total of 1449 patients were used for experiments and analysis in this retrospective study, who underwent spinal or abdominal CT scans for other indications between March 2018 and May 2020. All data was gathered from three different CT vendors. Among them, 586 cases were used for training, and other 863 cases were used for testing. A fully convolutional neural network, called U-Net, was employed for automated vertebral body segmentation. The manually sketched region of vertebral body was used as the ground truth for comparison. A convolutional neural network, called DenseNet-121, was applied for BMD calculation. The values post-processed by quantitative computed tomography (QCT) were identified as the standards for analysis. RESULTS: Based on the diversity of CT vendors, all testing cases were split into three testing cohorts: Test set 1 (n = 463), test set 2 (n = 200), and test set 3 (n = 200). Automated segmentation correlated well with manual segmentation regarding four lumbar vertebral bodies (L1-L4): the minimum average dice coefficients for three testing sets were 0.823, 0.786, and 0.782, respectively. For testing sets from different vendors, the average BMDs calculated by automated regression showed high correlation (r > 0.98) and agreement with those derived from QCT. CONCLUSIONS: A deep learning-based method could achieve fully automatic identification of osteoporosis, osteopenia, and normal bone mineral density in CT images. KEY POINTS: • Deep learning can perform accurate fully automated segmentation of lumbar vertebral body in CT images. • The average BMDs obtained by deep learning highly correlates with ones derived from QCT. • The deep learning-based method could be helpful for clinicians in opportunistic osteoporosis screening in spinal or abdominal CT scans.
Entities:
Keywords:
Bone density; Deep learning; Osteoporosis; Spine; Tomography, X-ray computed
Authors: Jennifer J Y Lee; Elaheh Aghdassi; Angela M Cheung; Stacey Morrison; Anne Cymet; Valentina Peeva; Carolyn Neville; Sara Hewitt; Deborah DaCosta; Christian Pineau; Janet Pope; Paul R Fortin Journal: J Rheumatol Date: 2012-06-01 Impact factor: 4.666
Authors: Kristine E Ensrud; Terri L Blackwell; Peggy M Cawthon; Douglas C Bauer; Howard A Fink; John T Schousboe; Dennis M Black; Eric S Orwoll; Deborah M Kado; Jane A Cauley; Dawn C Mackey Journal: J Bone Miner Res Date: 2015-08-03 Impact factor: 6.741
Authors: Paola Pisani; Maria Daniela Renna; Francesco Conversano; Ernesto Casciaro; Marco Di Paola; Eugenio Quarta; Maurizio Muratore; Sergio Casciaro Journal: World J Orthop Date: 2016-03-18
Authors: Ferdinand Roski; Johannes Hammel; Kai Mei; Thomas Baum; Jan S Kirschke; Alexis Laugerette; Felix K Kopp; Jannis Bodden; Daniela Pfeiffer; Franz Pfeiffer; Ernst J Rummeny; Peter B Noël; Alexandra S Gersing; Benedikt J Schwaiger Journal: Eur Radiol Date: 2019-05-21 Impact factor: 5.315
Authors: Perry J Pickhardt; Thang Nguyen; Alberto A Perez; Peter M Graffy; Samuel Jang; Ronald M Summers; John W Garrett Journal: Radiol Artif Intell Date: 2022-08-31