Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Deep learning takes the pain out of back breaking work - Automatic vertebral segmentation and attenuation measurement for osteoporosis

Schmidt, David LU ; Ulén, Johannes LU ; Enqvist, Olof LU ; Persson, Erik LU ; Trägårdh, Elin LU ; Leander, Peter LU and Edenbrandt, Lars LU (2022) In Clinical Imaging 81. p.54-59
Abstract

Background: Osteoporosis is an underdiagnosed and undertreated disease worldwide. Recent studies have highlighted the use of simple vertebral trabecular attenuation values for opportunistic osteoporosis screening. Meanwhile, machine learning has been used to accurately segment large parts of the human skeleton. Purpose: To evaluate a fully automated deep learning-based method for lumbar vertebral segmentation and measurement of vertebral volumetric trabecular attenuation values. Material and methods: A deep learning-based method for automated segmentation of bones was retrospectively applied to non-contrast CT scans of 1008 patients (mean age 57 years, 472 female, 536 male). Each vertebral segmentation was automatically reduced by 7 mm... (More)

Background: Osteoporosis is an underdiagnosed and undertreated disease worldwide. Recent studies have highlighted the use of simple vertebral trabecular attenuation values for opportunistic osteoporosis screening. Meanwhile, machine learning has been used to accurately segment large parts of the human skeleton. Purpose: To evaluate a fully automated deep learning-based method for lumbar vertebral segmentation and measurement of vertebral volumetric trabecular attenuation values. Material and methods: A deep learning-based method for automated segmentation of bones was retrospectively applied to non-contrast CT scans of 1008 patients (mean age 57 years, 472 female, 536 male). Each vertebral segmentation was automatically reduced by 7 mm in all directions in order to avoid cortical bone. The mean and median volumetric attenuation values from Th12 to L4 were obtained and plotted against patient age and sex. L1 values were further analyzed to facilitate comparison with previous studies. Results: The mean L1 attenuation values decreased linearly with age by −2.2 HU per year (age > 30, 95% CI: −2.4, −2.0, R2 = 0.3544). The mean L1 attenuation value of the entire population cohort was 140 HU ± 54. Conclusions: With results closely matching those of previous studies, we believe that our fully automated deep learning-based method can be used to obtain lumbar volumetric trabecular attenuation values which can be used for opportunistic screening of osteoporosis in patients undergoing CT scans for other reasons.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; and
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Artificial intelligence, Deep learning, Osteoporosis, Spiral computed, Tomography
in
Clinical Imaging
volume
81
pages
6 pages
publisher
Elsevier
external identifiers
  • pmid:34598006
  • scopus:85115938558
ISSN
0899-7071
DOI
10.1016/j.clinimag.2021.08.009
language
English
LU publication?
no
additional info
Publisher Copyright: © 2021 Elsevier Inc.
id
63843332-e083-4c9d-9633-02f5c87b7236
date added to LUP
2021-10-19 10:48:45
date last changed
2024-06-15 18:26:24
@article{63843332-e083-4c9d-9633-02f5c87b7236,
  abstract     = {{<p>Background: Osteoporosis is an underdiagnosed and undertreated disease worldwide. Recent studies have highlighted the use of simple vertebral trabecular attenuation values for opportunistic osteoporosis screening. Meanwhile, machine learning has been used to accurately segment large parts of the human skeleton. Purpose: To evaluate a fully automated deep learning-based method for lumbar vertebral segmentation and measurement of vertebral volumetric trabecular attenuation values. Material and methods: A deep learning-based method for automated segmentation of bones was retrospectively applied to non-contrast CT scans of 1008 patients (mean age 57 years, 472 female, 536 male). Each vertebral segmentation was automatically reduced by 7 mm in all directions in order to avoid cortical bone. The mean and median volumetric attenuation values from Th12 to L4 were obtained and plotted against patient age and sex. L1 values were further analyzed to facilitate comparison with previous studies. Results: The mean L1 attenuation values decreased linearly with age by −2.2 HU per year (age &gt; 30, 95% CI: −2.4, −2.0, R<sup>2</sup> = 0.3544). The mean L1 attenuation value of the entire population cohort was 140 HU ± 54. Conclusions: With results closely matching those of previous studies, we believe that our fully automated deep learning-based method can be used to obtain lumbar volumetric trabecular attenuation values which can be used for opportunistic screening of osteoporosis in patients undergoing CT scans for other reasons.</p>}},
  author       = {{Schmidt, David and Ulén, Johannes and Enqvist, Olof and Persson, Erik and Trägårdh, Elin and Leander, Peter and Edenbrandt, Lars}},
  issn         = {{0899-7071}},
  keywords     = {{Artificial intelligence; Deep learning; Osteoporosis; Spiral computed; Tomography}},
  language     = {{eng}},
  pages        = {{54--59}},
  publisher    = {{Elsevier}},
  series       = {{Clinical Imaging}},
  title        = {{Deep learning takes the pain out of back breaking work - Automatic vertebral segmentation and attenuation measurement for osteoporosis}},
  url          = {{http://dx.doi.org/10.1016/j.clinimag.2021.08.009}},
  doi          = {{10.1016/j.clinimag.2021.08.009}},
  volume       = {{81}},
  year         = {{2022}},
}