Application of ultrasound to measure coronal curvature and vertebral rotation in adolescent idiopathic scoliosis

  • Author / Creator
    Chen, Wei
  • Adolescent idiopathic scoliosis (AIS) is a three-dimensional (3D) spinal deformity occurring during adolescence with no known causes. Both the coronal curvature and vertebral axial rotation (VAR) are important parameters to assess the severity, predict the progression and evaluate the outcomes of AIS. Currently, 2D radiography is the standard imaging method to diagnose AIS and the Cobb angle is the clinical practice to measure the severity of AIS. However, the radiographic method exposes the patients to harmful ionizing radiation and has limitation to reveal the true nature of scoliosis. Ultrasound as a non-ionizing radiation method has been proposed in this PhD study to measure the coronal curvature and VAR in children with AIS. To optimize the ultrasound set up for spine imaging applications, experiments were performed to investigate the optimum configurations. Also, according to the ultrasound theory, the tissue bone interface can provide strong reflection signals when the surface is relatively flat. Therefore, after scanning a spinal phantom, the spinous processes (SP), laminae and transverse processes could be recognized on both the coronal and transverse views of the ultrasound images. Among these three landmarks, the center of lamina (COL) method was developed and reported to be the best estimation method for the coronal curvature measurement of scoliosis. The intra- and inter-observer reliabilities on both the in-vitro (a cadaver spine phantom) and in-vivo studies (26 Children with AIS) were found to be high (Intraclass correlation coefficient (ICC): >0.87, mean absolute difference (MAD): 1.3°-4.1°). In addition, the coronal curvature measurements using the COL method showed high agreement and small deviation (ICC: 0.92-0.96, MAD: 1.7°-2.9°) compared with the clinical record of the Cobb angle from the local scoliosis clinic which was obtained on the same day. To measure the VAR on the ultrasound transverse images, the COL method was used. In-vitro and in-vivo studies were then performed and the results demonstrated that the intra- and inter-observer reliabilities were high (ICC: 0.91-0.99, MAD: 0.3°-0.9°). The in-vitro study also showed that the VAR measurements from the ultrasound image using the COL method were more accurate than the VAR measurements from the radiographs using the Stokes method. To improve the quality of the image and reduce the human measurement errors on both the coronal curvature and VAR measurements using ultrasound, a semi-automatic measurement program was developed. The program includes three parts: the pre-processing part processed the original ultrasound data to improve the quality of the ultrasound images by reducing the sparkle noise using the wavelet soft threshold method and improve the processing time by reducing the data size; the image reconstruction part generated the coronal and transverse images for measurements; and the semi-automatic measurement part required the operators to point out the laminae, and then the program segmented the laminae from the background using optimum global thresholding based on the Otsu’s method, determined the centers of lamina more precisely, and automatically calculated the coronal curvature and the VAR measurements. The reliability and validity of the measurements were investigated on the in-vitro and in-vivo data. The coronal curvature measurements had high intra-observer reliability (ICC: 0.85-0.98, MAD: 1.4°-2.4°). However, only fair inter-observer reliability was obtained for the in-vivo data (ICC: 0.76, MAD: 3.4°). The VAR measurements by the program showed high intra- and inter-observer reliabilities (ICC: >0.94, MAD: 0.2°-0.9°). This thesis reported that 1) the ultrasound imaging method could be used to assess the coronal curvature and VAR of AIS; 2) a new center of lamina (COL) method was developed and validated to measure the coronal curvature and VAR on ultrasound images reliably; 3) a semi-automatic program was developed to improve image quality and reduce the human measurement errors for both of the coronal curvature and VAR measurements.

  • Subjects / Keywords
  • Graduation date
  • Type of Item
  • Degree
    Doctor of Philosophy
  • DOI
  • License
    This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.
  • Language
  • Institution
    University of Alberta
  • Degree level
  • Department
    • Department of Biomedical Engineering
  • Supervisor / co-supervisor and their department(s)
    • Le, Lawrence (Radiology and Diagnostic Imaging)
    • Lou, Edmond (Surgery, Electrical and Computer Engineering)
  • Examining committee members and their departments
    • Adeeb, Samer (Civil and Environmental Engineering)
    • Wilman, Alan (Biomedical Engineering)
    • Parent, Eric (Physical Therapy)
    • Vette, Albert (Mechanical Engineering)
    • Villemure, Isabelle (Mechanical Engineering)