Study group
A total of 121 volunteers (56 females; 65 males) were included in this study. Inclusion criteria were age over 18 years, ability to walk on a treadmill for a minimum of 90 seconds with a maximum speed of 5 km/h, as well as a body mass index of under 35 kg/m2. Excluded from this study were participants with musculoskeletal disorders, such as osteoarthritis and rheumatoid arthritis, any history of previous fractures as well as systemic diseases that could potentially influence walking on a treadmill. All participants gave their oral and written consent to participate in this study and were given the option to discontinue their participation at any time. The local human ethics committee of the institute approved the protocol of the study (study number EK 3953).
Rasterstereographic system
The “Formetric 4D Motion®“ system (DIERS International GmbH, Germany) is based on the method of stereophotogrammetric surface measurements of the back (8). It uses a slide projector to project horizontal parallel light lines onto the back surface of patients. Then an image or video is taken of the back surface with a digital network camera at 50 Hz. This camera uses a CMOS sensor with a resolution of 1280x1024 pixels. The software performs a digital reconstruction of the back surface by transforming the horizontal light lines and their corresponding curvatures into a three-dimensional scatter plot. Based on the specific shape of the spinous processes of the vertebra prominence (VP) and the concavity of the lumbar dimples, as anatomical reference points, a model of the spine is created. With this technique transverse, axial and sagittal profiles as well as several spinal and pelvic angles can be calculated. In contrast to other surface topography systems, rasterstereography allows an analysis not only of the back surface, but also of the underlying spine. This is made possible by using a spine model created by Drerup and Turner-Smith, based on more than 500 whole spine radiographs of patients with scoliosis (8, 10). Due to movements of the skin and soft tissues above the anatomical landmarks during motion it is necessary to use infrared reflecting markers on the VP and the two lumbar dimples. In a previous study, Betsch et al. were able to demonstrate that the dynamic rastersterographic system is able to detect these markers with an accuracy of +/- 1mm under static and dynamic conditions using an array of 8 LEDs (8). The position of the markers can be automatically detected with the use of an algorithm that scans the back surface for all bright elliptical regions. Then from the position of the markers a sub-pixel approximation of the centre of the marker is used to determine its exact position. In approximately 100ms a complete reconstruction of the back surface is possible with these algorithms, making a real-time display of the spine and back surface during dynamic measurements possible.
Measuring setup
All volunteers were measured in shorts and women were offered a drape to cover their chest so that it would not be visible from the back. Care was taken to provide free views on the VP and the two lumbar dimples. Three flat adhesive markers were placed on the back surface of the volunteers, one on the VP and two on the lumbar dimples, by the examiner. To verify the correct marker position, we performed a static scan of the back surface, and if necessary the position of the marker was corrected. Once the correct placement of the marker was confirmed, they were kept in place for the following dynamic measurements.
Measuring protocol
After confirmation of the correct marker placement a total of three static measurements were conducted. For all static measurements, the volunteers were placed on a treadmill with a distance of 2 metres from the camera. All volunteers were instructed to stand in the neutral standing position with their arms hanging to the sides and extended knees. All dynamic measurements were conducted while the volunteers were walking on a treadmill to ensure an approximate distance to the camera of about 2 metres. The subjects were measured three times while walking with velocities of 1, 2, 4 and 5 km/h. On average, all volunteers walked for 60 seconds on the treadmill to adjust to the treadmill velocity. The subsequent dynamic rasterstereographic recording lasted for 6 seconds, which resulted in 330 to 335 frames per measurement. After the recording, the treadmill was automatically stopped by the system and the software processed the collected data, while the subjects were resting for 2 minutes before the next trial.
Data analysis
All parameters analyzed in this study are defined in table 2. The average values of all three static and dynamic trials, including all 4 different velocities, were calculated by the examiner. Based on these values we calculated the standard deviations for all parameters and conditions to be able to compare static with dynamic and all dynamic measurements.
Statistical analysis
All data were processed and prepared by the programme “Dicam II 2.2.3” (DIERS International GmbH, Germany). SPSS Statistics Version 20 was used for all statistical analysis (IBM, USA). The differences in the anthropometric data and static parameters between male and female participants were evaluated with the Mann-Whitney U Test (level of significance α < 0,05). As the Kolmogorov-Smirnov-Test revealed no normal distribution for any data in this study, we used non-parametric statistical tests. A Kruskal-Wallis-Test was used to analyse the parameters collected during different walking speeds for significant differences. If so, significance was verified by using a Mann-Whitney U-Test. Furthermore, the level of significance was corrected by using Bonferroni correction at a significance level of p< 0,0083 for analysing data during different walking speeds. For comparing dynamic with static parameters, the same procedure was used, with a Bonferroni correction at a significance level of p<0,0125.