A Deep Learning Model for Three-Dimensional Determination of Whole Thoracic Vertebral Bone Mineral Density from Noncontrast Chest CT: The Multi-Ethnic Study of Atherosclerosis.

Journal: Radiology
PMID: 40067103

Abstract

Background Recent studies have investigated how deep learning (DL) algorithms applied to CT using two-dimensional (2D) segmentation (sagittal or axial planes) can calculate bone mineral density (BMD) and predict osteoporosis-related outcomes. Purpose To determine whether TotalSegmentator, an nnU-net algorithm, can measure three-dimensional (3D) vertebral body BMD across consistently imaged thoracic levels (T1-T10) at any conventional, noncontrast chest CT examination. Materials and Methods This study is a secondary analysis of a multicenter ( = 6) prospective cohort, the Multi-Ethnic Study of Atherosclerosis (MESA). Participants underwent noncontrast chest CT with ( = 296) and without ( = 2660) a phantom. In 594 participants, manual segmentation for T1-T10 vertebrae was performed on axial and sagittal planes. TotalSegmentator provided 3D vertebral body segmentation of T1-T10 levels with further postprocessing to remove cortical bone. Two-dimensional axial and sagittal DL-derived algorithms were developed and compared with 3D model performance. Dice and intersection-over-union scores were calculated. Vertebral BMD-derived data, integrated with the Fracture Risk Assessment Tool with no BMD (FRAXnb), were used to predict incident vertebral fractures (VFx) in participants from the follow-up MESA Examination 6 ( = 1304). Results This study included 2956 participants (1546 [52%] female; age, 69 years ± 9 [SD]), with longitudinal data obtained approximately 6.2 years later in a subset of 1304 participants. DL-derived 3D segmentations were correlated with manual axial (Dice score, 0.93; 95% CI: 0.92, 0.95) and sagittal (Dice score, 0.91; 95% CI: 0.88, 0.93) segmentations. DL-derived 2D axial and sagittal BMD measurements had higher uncertainty compared with DL-derived 3D BMD measurements (average SDs, 2D axial and 2D sagittal vs 3D BMD: 65 mg/cm and 59 mg/cm vs 41 mg/cm, respectively; both < .001). Three-dimensional vertebral BMD with FRAXnb demonstrated better performance in predicting incident VFx (area under the receiver operating characteristic curve [AUC], 0.82) compared with FRAXnb alone (AUC, 0.66; = .03). Conclusion A multilevel DL algorithm for measuring 3D whole thoracic vertebral BMD using conventional chest CT determined distinct BMD patterns from whole thoracic vertebrae and provided incremental value in predicting VFx. ClinicalTrials.gov identifier: NCT00005487 © RSNA, 2025 . See also the editorial by Steiger in this issue.

Authors

  • Quincy A Hathaway
    Division of Exercise Physiology, West Virginia University School of Medicine, PO Box 9227, 1 Medical Center Drive, Morgantown, WV, 26505, USA.
  • Arta Kasaeian
    Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Russell H. Morgan, Baltimore, MD, USA.
  • Tommy Pan
    Department of Radiology and Radiologic Sciences, Johns Hopkins University, Baltimore, Md.
  • David A Bluemke
    From the Department of Radiology (B.A.-V.), Bloomberg School of Public Health (E.G.), and Department of Medicine, Cardiology and Radiology (J.A.C.L.), Johns Hopkins University, Baltimore, MD; George Washington University, DC (X.Y.); Office of Biostatistics, NHLBI, NIH, Bethesda, MD (C.O.W.); Department of Preventive Medicine, Northwestern University Medical School, Chicago, IL (K.L.); Department of Cardiology, Wake Forest University Health Sciences, Winston-Salem, NC (W.G.H.); Department of Biostatistics, University of Washington, Seattle (R.M.); Department of Radiology, UCLA School of Medicine, Los Angeles, CA (A.S.G.); Division of Epidemiology and Community Health, University of Minnesota, Minneapolis (A.R.F.); Departments of Medicine and Epidemiology, Columbia University, New York, NY (S.S.); and Radiology and Imaging Sciences, NIH Clinical Center, Bethesda, MD (D.A.B.).
  • Elena Ghotbi
    From the Russell H. Morgan Department of Radiology and Radiological Science (S.D., J.G.K., E.G., H.A.I., K.M., K.T., E.K.F.) and Department of Biomedical Engineering (W.B.Z.), Johns Hopkins University School of Medicine, 601 N Carolina St, Baltimore, MD 21287; Division of Musculoskeletal Imaging, Department of Radiology, Mayo Clinic, Rochester, Minn (F.I.B.); Department of Radiology, New York University Grossman School of Medicine, New York, NY (J.F.); Department of Radiology and Imaging, Hospital for Special Surgery, New York, NY (J.A.C.); and Department of Radiology, Quantitative Imaging Center, Boston University School of Medicine, Boston, Mass (A.G.).
  • Joshua G Klein
    From the Russell H. Morgan Department of Radiology and Radiological Science (S.D., J.G.K., E.G., H.A.I., K.M., K.T., E.K.F.) and Department of Biomedical Engineering (W.B.Z.), Johns Hopkins University School of Medicine, 601 N Carolina St, Baltimore, MD 21287; Division of Musculoskeletal Imaging, Department of Radiology, Mayo Clinic, Rochester, Minn (F.I.B.); Department of Radiology, New York University Grossman School of Medicine, New York, NY (J.F.); Department of Radiology and Imaging, Hospital for Special Surgery, New York, NY (J.A.C.); and Department of Radiology, Quantitative Imaging Center, Boston University School of Medicine, Boston, Mass (A.G.).
  • Hamza Ahmed Ibad
    From the Russell H. Morgan Department of Radiology and Radiological Science (S.D., J.G.K., E.G., H.A.I., K.M., K.T., E.K.F.) and Department of Biomedical Engineering (W.B.Z.), Johns Hopkins University School of Medicine, 601 N Carolina St, Baltimore, MD 21287; Division of Musculoskeletal Imaging, Department of Radiology, Mayo Clinic, Rochester, Minn (F.I.B.); Department of Radiology, New York University Grossman School of Medicine, New York, NY (J.F.); Department of Radiology and Imaging, Hospital for Special Surgery, New York, NY (J.A.C.); and Department of Radiology, Quantitative Imaging Center, Boston University School of Medicine, Boston, Mass (A.G.).
  • Chris Dailing
    Lundquist Institute at Harbor-University of California Los Angeles School of Medicine, Torrance, Calif.
  • Geoffrey H Tison
    Department of Medicine (G.H.T., M.H.L., E.F., M.A.A., C.J., K.E.F., R.C.D.).
  • R Graham Barr
    Columbia University, Division of General Medicine, New York, NY, USA.
  • Wendy Post
    Division of Cardiology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Md.
  • Matthew Allison
    Department of Family Medicine and Public Health, University of California San Diego, La Jolla, CA, USA.
  • João A C Lima
    From the Department of Radiology (B.A.-V.), Bloomberg School of Public Health (E.G.), and Department of Medicine, Cardiology and Radiology (J.A.C.L.), Johns Hopkins University, Baltimore, MD; George Washington University, DC (X.Y.); Office of Biostatistics, NHLBI, NIH, Bethesda, MD (C.O.W.); Department of Preventive Medicine, Northwestern University Medical School, Chicago, IL (K.L.); Department of Cardiology, Wake Forest University Health Sciences, Winston-Salem, NC (W.G.H.); Department of Biostatistics, University of Washington, Seattle (R.M.); Department of Radiology, UCLA School of Medicine, Los Angeles, CA (A.S.G.); Division of Epidemiology and Community Health, University of Minnesota, Minneapolis (A.R.F.); Departments of Medicine and Epidemiology, Columbia University, New York, NY (S.S.); and Radiology and Imaging Sciences, NIH Clinical Center, Bethesda, MD (D.A.B.). jlima@jhmi.edu.
  • Matthew Budoff
    4 Division of Cardiology Los Angeles Biomedical Research at Harbor-UCLA Medical Center Torrance CA.
  • Shadpour Demehri
    Department of Radiology, The Johns Hopkins Hospital, Baltimore, MD 21287.

Keywords

External Resources

Popular Topics
Recent Journals