Abstract
Introduction
Diagnosis of osteoarthritis relies primarily on image-based analyses. X-ray, CT, and MRI can be used to evaluate various features associated with OA including joint space narrowing, deformity, articular cartilage integrity, and other joint parameters. While effective, these exams are costly, may expose the patient to ionizing radiation, and are often conducted under passive, non-weightbearing conditions. A supplemental form of analysis utilizing vibroarthrographic (VAG) signals provides an alternative that is safer and more cost-effective for the patient. The objective of this study is to correlate the kinematic patterns of normal, diseased (pre-operative), and implanted (post-operative) hip subjects to their VAG signals that were collected and to more specifically, determine if a correlation exists between femoral head center displacement and vibration signal features.
Methods
Of the 28 hips that were evaluated, 10 were normal, 10 were diseased, and 8 were implanted. To collect the VAG signal from each subject, two uniaxial accelerometers were placed on bony landmarks near the joint; one was placed on the greater trochanter of the femur and the other along the anterior edge of the iliac crest. The subjects performed a single cycle gait (stance and swing phase) activity under fluoroscopic surveillance. The CAD models of the implanted components were supplied by the sponsoring company while the subject bone models were created from CT scans. 3D-to-2D registration was conducted on subject fluoroscopic images to obtain kinematics, contact area, and femoral center head displacement. The VAG signals were trimmed to time, passed with a denoise filter and wavelet decomposition.
Results
When comparing the femoral head displacement to the vibration signals with respect to the normal hips, insignificant magnitudes of vibration were present (0.05 volts). For the diseased hips, greater magnitudes were seen (0.2 volts). For the implanted subjects, the overall vibration features were small (0.05 volts) much like the signals from the normal hips except for spikes that correlated to features within the gait cycle. Therefore, grinding sounds were heard from the degenerative hips, but not present for the normal or implanted hips in this study.
Discussion
In regards to the normal hip subjects, the lesser magnitude of volts correlated well with the kinematic results showing no separation of the femoral head center (1 mm). For the diseased hips, the instances of greater feature quantity occurred at moments where the subjects experienced higher values of head center displacement (1 mm). These subjects also had an overall increase in average voltage magnitude likely due to the loss of cartilage about the articulating surface resulting in a rougher surface for the accelerometers to record. For the implanted subjects, due to no head center displacement and a smoother surface for joint articulation, the vibration signals were smaller than the diseased case but showed better correlation with features within the gait cycle. No exact quantification has been determined between separation and accelerometer voltage output, further studies and testing will need to be carried out in order to reach such a conclusion.
For any figures or tables, please contact authors directly.
