Sacrum

Twenty-eight healthy subjects (15 females, 13 males; age 24 ± 2.70 years; 70 ± 12.70 kg and 1.76 ± 0.09 m in height) participated in the study. Each of the subjects performed two test sessions on two days (6.75 ± 2.26 days in between). A test session consisted of one static neutral zero position (n-pose) sequence and a 6 min walk test [24 (link)]. The study was approved by the ethical committee of the Technische Universität Kaiserslautern (TUK) and meets the criteria of the declaration of Helsinki. After receiving all relevant study information, the participants signed an informed consent to the study including a permission to publish data.
On both test days lower extremity 3D kinematics was simultaneously captured using twelve OptiTrack Prime 13 cameras (NaturalPoint, Inc., Corvallis, OR, USA) and seven XSens MTw Awinda (Xsens Technologies BV, Enschede, The Netherlands) IMUs.
IMUs were activated at least 20 min before measurement start. A static trial was performed before each subject was instrumented, with the sensors lying still for a period of at least 10 s, to estimate and subtract the gyroscope bias. These steps were conducted in accordance with the recommendations of Bergamini et al. [21 (link)].
Thirty-two retroreflective markers were attached to anatomical landmarks (AL) according to Leardini et al. [25 (link)] and the OptiTrack recommendations. Each IMU was secured in matched 3D printed boxes to which four markers were rigidly attached. These markers were used for unique identification in the optical point cloud as well as for orientation estimation. Using the OptiTrack Software, the origin of the boxes was moved to the center of the attached sensor casing. These box/sensor compounds were fixed to the body segments using straps and double-sided adhesive tape. IMUs were attached on the right and left dorsum of the foot approximately atop the base of metatarsal II-IV, on the right and left lateral aspect of the shank, due to better visibility, on the right and left lateral aspect of the lower third of the thigh and between the Spinae Iliacae Posteriores Superiores approximately atop the sacral base (Figure 1).
Inertial and optical data were simultaneously recorded at 60 Hz using XSens MVN Biomech (Version 4.3.7) and OptiTrack Motive (Version 1.10.0) which were hardware synchronized using a standard 5 V TTL signal. The alignment orientations between the IMUs and the rigid boxes were calculated using the method described in [7 (link)]. The biomechanical model according to Cappozzo et al. [26 (link)] and the IMU-to-segment calibrations were extracted from the OMC data of the n-pose sequence. The joint centers were also calculated from the OMC data during the n-pose sequence according to the definitions of Visual3D (C-Motion, Inc, Germantown, MD, USA), a widely used software tool for 3D biomechanics research. The first OMC frame of each walking sequence was used as initialization for the IMU-based kinematics estimation. Both systems used the same biomechanical model.
The inertial data was processed with an iterated extended Kalman filter (IEKF) approach based on [14 ] while omitting magnetometer information. The gyroscope biases were extracted from a static sequence (see above), while the accelerometer biases were estimated in the IEKF along with the kinematics estimation using the model described in [11 (link),27 ]. The same sequence was processed twice: initially to obtain a converged estimate of the acceleration bias, which was then used as initial guess in the second run. The estimated segment orientations were used to derive relative joint orientations. These were decomposed into joint angles using Euler angle decomposition [28 (link)]. The sensor fusion method is detailed in Appendix A.
To minimize STA, the OMC-based joint angles were derived from marker clusters on the rigid boxes (condition 1). For secondary analyses the joint angles were calculated based on the markers attached to the anatomical landmarks (condition 2). Initial contact (IC) was detected based on the left and right heel marker [29 (link)]. Turning phases in the 6 min walk tests were omitted. In order to investigate drift behavior, 10 left and right steps (one trial) were identified at three sections, i.e., beginning (A), middle (B), and end (C) of the test. All joint angle curves were normalized to 100 percent gait cycle (GC).

For all 12 models, we selected and extracted the meshes of each vertebra from T2 to the sacrum. The total stress of the vertebral mesh was added and equalized according to the volume of each vertebra. We collected data on stress from the screws at each vertebra. The rod was extracted separately, and the von Mises stress was evaluated. We analyzed the von Mises stress on each vertebra and implant, focusing on the differences in spinal balance, fusion length, and implants in UIV. Previous studies have validated the use of FEM analysis for examining von Mises stress in the lumbar vertebrae [6 (link)–8 (link)].