Measurement of standing whole spine alignment with the EOS system
Comparison of whole spine alignment between the supine and standing positions necessitates high reproducibility of the imaging technology. Conventional X-ray measurement with a cone-beam X-ray provides significant magnification of the subject within the margin of the cassette. Therefore, we used the EOS system (EOS Imaging, Paris, France) as a principal device for measuring whole spine alignment [16]. The EOS system is a slot-scanning three-dimensional (3D) X-ray imager developed in collaboration by multidisciplinary partners, radiation physics engineers, biomechanical engineers, medical radiologists, and orthopedic pediatric surgeons to overcome the limitations of conventional X-ray measurement. From the simultaneous anteroposterior and lateral X-rays of the whole body to the 3D bone external envelope technique, 3D reconstruction is possible at every level of the osteoarticular system and especially the spine in the standing position. The EOS allows for more precise bone reconstruction in orthopedics, especially at the level of the spine, pelvis, and lower limbs, with limited X-ray exposure compared with conventional X-rays and CT scans [6, 7, 9, 19].
Radiographs with the EOS system [3, 9] were routinely obtained, as follows: 1) EOS radiographs were made from the head, including the center of the auditory canal, to the feet. 2) Each patient was asked to stand comfortably on a force plate while placing their hands on their cheeks. 3) A mirror placed at eye level in the inner wall of the EOS cabin helped the patient maintain a horizontal gaze [16].
The default scan speed of the EOS system was 7.6 cm/s. Acquisition time was linked to scan height: Time of acquisition (s) = height of acquisition (cm)/7.6. Scan speed can be increased if the patient is restless and having difficulty keeping still during the acquisition. Nevertheless, subtle artifacts in the images can occur due to body sway during scanning, but these are minimized because of the rapid X-ray detection time (0.8333 ms) with no blurring of the images. Some accessories are available to stabilize the patient in the EOS cabin. A recent study demonstrated that motion artifacts do not affect spinal measurements [26].
Measurement of supine whole spine alignment with CT-generated digital reconstructed radiographs
Following EOS scanning, the whole spine, including the head and pelvis of the patient, was also scanned in the supine position with CT (Activion16, TSX-031A, Toshiba Medical Systems Corp., Tochigi, Japan). Currently, the most accurate 3D bone information may be obtained with CT imaging. To eliminate software-related bias and to guarantee equivalent computational methods during spinopelvic parameter comparisons between the supine position in the CT and the standing position in the EOS, the same analysis software must be used. Therefore, we chose to transform the CT dataset into an EOS-like dataset using the digitally reconstructed radiograph (DRR) technique. The DRR technique consists of simulating X-rays passing through the reconstructed CT volume based on an absorption-only optical model, thus generating an X-ray-like image. These biplanar projections were reconstructed using the same calibration parameters and geometry as the EOS cabin (Fig. 1). These projected anteroposterior and lateral DRRs, as well as those performed with the EOS system, were used as inputs for stereoradiographic spine modeling. Thus, for each position, 3D spine modeling was obtained using sterEOS software (sterEOS 1.6, EOS Imaging, Paris, France) [15] with both the EOS data (standing position) and the CT-generated DRRs (supine position), and compared between the two positions (Fig. 2).
Spino-pelvic parameters for comparison
Using the full-spine workflow of the sterEOS software, the following spinopelvic parameters were calculated: kyphosis T1–T12 and T4–12, lumbar lordosis (LL) with respect to L1–L5 and L1–S1, pelvic tilt (PT), sacral slope (SS), and pelvic incidence (PI). PI is the angle between the line perpendicular to the sacral plate at its midpoint and the line connecting the midpoint of the sacral plate to the center of the axis connecting both acetabulae. PT is the angle defined between the line connecting the midpoint of the sacral plate to the center of the axis of both acetabulae and the vertical axis. SS is the angle between the sacral plate and the horizontal line. Regarding the deformity, the Cobb angle of the major curve (Cobb angle) and the axial vertebral rotation of the apex in the major curve (Rotation) were measured. All of the parameter results were compared between the supine (CT-generated DRR) and standing (EOS) positions.
Clinical subjects
After obtaining institutional review board approval (Approval number 2, 27th Dec., 2013, Institutional Review Board of Kameda Daiichi Hospital, Niigata, Japan), we prospectively enrolled patients with ASD under the following criteria: 1) diagnosis: degenerative and idiopathic spinal kyphoscoliosis with Cobb angle more than 30° or degenerative kyphosis with PI-LL mismatch more than 20°, 2) age: > 20 years old, 3) sex: woman, 4) candidate for surgical treatment, 5) a full-spine CT scan (acquired from auditory canals to the proximal third of the femur) and a full-spine EOS image in both preoperative and postoperative states, 6) study term: from April 2014 to March 2016. The following demographic characteristics were obtained for each patient: age, sex, weight, and height. The body mass index was calculated as the weight in kilograms divided by the square of the height in meters. Patients with transitional vertebrae were excluded for precise measurement and comparison between supine and standing positions. Regarding the gender difference, we found a significant difference between men and women in PT, pelvic thickness, SVA, and lower extremity alignment in the previous study [17]. If men are included, the result may be affected. Therefore, men are excluded in this study. Consequently, a total of 24 cases with a mean age of 60.1 years (range: 20–80 years; 24 women) were analyzed after obtaining of informed consent for participation in the study. We used the Japanese version of the Oswestry Disability Index (ODI) [12, 13] and Scoliosis Research Society-22 score (SRS-22) [2, 18] to assess the health-related quality of life. ODI and SRS-22 are the principal condition-specific outcome measures used in the management of low back disorders and spinal deformities, respectively. Normal values without symptoms are 0 (%) in the ODI and 5 in the SRS-22, with the worst values being 100 (%) in the ODI and 0 in the SRS-22.
Statistical analysis
JMP (version 9; SAS Institute, Cary, NC) and SPSS (IBM SPSS Statistics for Windows, Version 24.0, IBM Corp., Armonk, NY) were used for all statistical analyses. Mean, range, standard deviation (SD), standard error (SE), and the interquartile range, 25%/75%, were calculated for all the demographic and radiographic parameters. All variance-dependent variables were checked for normality and homogeneity of variance. Alpha was set at p < 0.05.
An intra-class correlation coefficient (ICC) was calculated to explore consistency within and between examiners for measurements with the original EOS images and the CT-generated DRRs. To evaluate intra-rater reliability, we compared the measurements obtained by two examiners who completed the EOS measurement training and had worked with these measurements for 3 years, and measured all the parameters of the 24 subjects twice with a 1-week interval. To evaluate inter-rater reliability, we compared the measurements obtained by the two examiners of all the parameters of the same 24 subjects in 1 week. An ICC value approaching 1.0 indicates less variability, better consistency, and a value over 0.8 is considered sufficiently reliable.
The values of all the alignment spinopelvic parameters were normally distributed, thus a paired-t-test was performed to compare between supine (CT-generated DRR) and standing (EOS) positions. Type I error (α), power (1-β), and post-hoc sample size for the statistical significance were calculated.