A systematic approach to feature tracking of lumbar spine vertebrae from fluoroscopic images using complex-valued wavelets

Digital tracking algorithm reveals the influence of structural irregularities on joint movements in the human cervical spine

BACKGROUND

Disc height loss and osteophytes change the local mechanical environment in the spine; while previous research has examined kinematic dysfunction under degenerative change, none has looked at the influence of disc height loss and osteophytes throughout movement.

METHODS

Twenty patients with pain related to the head, neck or shoulders were imaged via videofluoroscopy as they underwent sagittal-plane flexion and extension. A clinician graded disc height loss and osteophytes as "severe/moderate", "mild", or "none". A novel tracking algorithm quantified motions of each vertebra. This information was used to calculate intervertebral angular and shear displacements. The digital algorithm made it practical to track individual vertebrae in multiple patients through hundreds of images without bias.

FINDINGS

Cases without height loss/osteophytes had a consistent increase in intervertebral angular displacement from C2/C3 to C5/C6, like that of healthy individuals, and mild height losses did not produce aberrations that were systematic or necessarily discernable. However, joints with moderate to severe disc height loss and osteophytes exhibited reduced range of motion compared to adjacent unaffected joints in that patient and corresponding joints in patients without structural irregularities.

INTERPRETATION

Digitally-obtained motion histories of individual joints allowed anatomical joint changes to be linked with changes in joint movement patterns. Specifically, disc height loss and osteophytes were found to influence cervical spine movement in the sagittal plane, reducing angular motions at affected joints by approximately 10% between those with and without height loss and osteophytes. Further, these joint changes were associated with perturbed intervertebral angular and shear movements.

A videofluoroscopy-based tracking algorithm for quantifying the time course of human intervertebral displacements

The motions of individual intervertebral joints can affect spine motion, injury risk, deterioration, pain, treatment strategies, and clinical outcomes. Since standard kinematic methods do not provide precise time-course details about individual vertebrae and intervertebral motions, information that could be useful for scientific advancement and clinical assessment, we developed an iterative template matching algorithm to obtain this data from videofluoroscopy images. To assess the bias of our approach, vertebrae in an intact porcine spine were tracked and compared to the motions of high-contrast markers. To estimate precision under clinical conditions, motions of three human cervical spines were tracked independently ten times and vertebral and intervertebral motions associated with individual trials were compared to corresponding averages. Both tests produced errors in intervertebral angular and shear displacements no greater than 0.4° and 0.055 mm, respectively. When applied to two patient cases, aberrant intervertebral motions in the cervical spine were typically found to correlate with patient-specific anatomical features such as disc height loss and osteophytes. The case studies suggest that intervertebral kinematic time-course data could have value in clinical assessments, lead to broader understanding of how specific anatomical features influence joint motions, and in due course inform clinical treatments.