Description of soft tissue artifacts and related consequences on hindlimb kinematics during canine gait

Background Soft tissue artifacts (STAs) are a source of error in marker-based gait analysis in dogs. While some studies have revealed the existence of STAs in the canine hindlimb, STAs and their influence on kinematic gait analysis remain unclear. Methods Thirteen healthy Taiwan dogs affixed with twenty skin markers on the thigh and crus were recruited. Soft tissue artifacts and their influence on the determination of segment poses and stifle angles were assessed by simultaneously measuring marker trajectories and kinematics of the underlying bones via a model-based fluoroscopic analysis method. Results Markers on the thigh showed higher STAs than those on the crus, with root-mean-square amplitudes up to 15.5 mm. None of the tested marker clusters were able to accurately reproduce the skeletal poses, in which the maximum root-mean-square deviations ranged from 3.4° to 8.1°. The use of markers resulted in overestimated stifle flexion during 40–60% of the gait cycle and underestimated stifle flexion during 80–90% of the gait cycle. Conclusions Considerable magnitudes and effects of STAs on the marker-based 3D gait analysis of dogs were demonstrated. The results indicate that the development of error-compensation techniques based on knowledge regarding STAs is warranted for more accurate gait analysis.

Multisegment Kinematics of the Spinal Column: Soft Tissue Artifacts Assessment

A major challenge in the assessment of intersegmental spinal column angles during trunk motion is the inherent error in recording the movement of bony anatomical landmarks caused by soft tissue artifacts (STAs). This study aims to perform an uncertainty analysis and estimate the typical errors induced by STA into the intersegmental angles of a multisegment spinal column model during trunk bending in different directions by modeling the relative displacement between skin-mounted markers and actual bony landmarks during trunk bending. First, we modeled the maximum displacement of markers relative to the bony landmarks with a multivariate Gaussian distribution. In order to estimate the distribution parameters, we measured these relative displacements on five subjects at maximum trunk bending posture. Then, in order to model the error depending on trunk bending angle, we assumed that the error grows linearly as a function of the bending angle. Second, we applied our error model to the trunk motion measurement of 11 subjects to estimate the corrected trajectories of the bony landmarks and investigate the errors induced into the intersegmental angles of a multisegment spinal column model. For this purpose, the trunk was modeled as a seven-segment rigid-body system described using 23 reflective markers placed on various bony landmarks of the spinal column. Eleven seated subjects performed trunk bending in five directions and the three-dimensional (3D) intersegmental angles during trunk bending were calculated before and after error correction. While STA minimally affected the intersegmental angles in the sagittal plane (<16%), it considerably corrupted the intersegmental angles in the coronal (error ranged from 59% to 551%) and transverse (up to 161%) planes. Therefore, we recommend using the proposed error suppression technique for STA-induced error compensation as a tool to achieve more accurate spinal column kinematics measurements. Particularly, for intersegmental rotations in the coronal and transverse planes that have small range and are highly sensitive to measurement errors, the proposed technique makes the measurement more appropriate for use in clinical decision-making processes.