Accuracy of the Peak (TM) Two and Three-Dimensional Videography Analysis for a Rearfoot Model
Abstract
The purpose of this study was to determine the accuracy of the Peak 2D and 3 D video analysis systems for measuring static model angles. A model. representing the right rearfoot. was constructed from a squared and leveled piece of wood and a goniometer. The goniometer was rotated clock-wise or counter-clockwise in the frontal plane to simulate eversion or inversion of the shank relative to the calcaneus. The entire model was also rotated clockwise or counter-clockwise in either the transverse or sagittal planes. Three video cameras, two for 3D analysis and one for 2D analysis, simultaneously recorded fifty-one different positions of the rearfoot model. Special attention was given to field size (0.75 meters) and distance between markers for recording the 51 conditions. A repeated measures ANOVA and Tukey's HSD detected significant difference of the inversion/eversion angle between methods: Actual angles, 2D joint angles, and 3D Segmental angles. There was a significant difference between Actual angle and 2D Joint angle and 2D Joint and 3D Segmental angle. The greatest absolute differences for 2D Joint angles occurred when the model was rotated 30 degrees in the transverse plane. These 12 conditions of 30 degree transverse rotations were significantly different than the other 39 conditions. Separate ANOVA's were applied to each group (n 12, n = 39). 3D was significantly better than 2D for the 12 conditions of 30 degree transverse plane rotations, however there was no difference among 2D and 3D values for all other conditions (n = 39). For the 51 conditions, the mean absolute error between Actual and 2D was 0.74 and Actual and 3D was 0.81 degrees.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1993
- Accession Number
- ADA267805
Entities
People
- Andrew M. Glaves
Organizations
- Air Force Institute of Technology