Abstract
Introduction
Kinematics analyses of the spine have been recognized as an effective method for functional analysis of the spine. CT is suitable for obtaining bony geometry of the vertebrae but radiation is a clinical concern. MRI is noninvasive but it is difficult to detect bone edges especially at endplates and processes where soft tissues attach. Kinematics analyses require tracking of solid bodies; therefore, bony geometry is not always necessary for kinematics analysis of the spine. This study aimed to develop a reliable and robust method for kinematics analysis of the spine using an innovative MRI-based 3D bone-marrow model.
Materials and Methods
This IRB-approved study recruited 17 patients undergoing lumbar decompression surgery to treat a single-level symptomatic herniation as part of a clinical trial for a new dynamic stabilization device. T1 & T2 sagittal MRI scans were acquired as part of the pre-operative evaluation in three positions: supine and with the shoulders rotated 45° to the left and right to induce torsion of the lumbar spine. 3D bone-marrow models of L5 and S1 at the neutral and rotated positions were created by selecting a threshold level of the bone-marrow intensity at bone-marrow/bone interface. Validated 3D-3D registration techniques were used to track movements of L5 and S1. Segmental movements at L5/S1 during torsion were calculated.
Results
Bone-marrow models were created not only in the vertebral body but also in superior/inferior, transverse and spinous processes, pedicles and laminae. Segmental rotation (mean±SD) at L5/S1 was shown to be symmetric for both left and right motions (p=0.149; Left: 1.04°±0.93° and Right: 1.33°±0.80°). The range of motion recorded was: left [0.05°-3.70°] and right [0.35°-3.25°]. These values were equivalent to previously reported values of axial lumbar rotation measured by 3D CT lumbar models.
Conclusions
This study demonstrated feasibility of kinematic analyses using the 3D bone-marrow model created with clinical MRI. The bone-marrow model shows the bone-marrow/bone interface geometry –the internal structure of the vertebra rather than outside geometry usually used for kinematic analyses– that is easily and consistently detected due to its high-contrast interface MRI intensity, which does not require lengthy manual tracing of the bony contour. The bone-marrow model includes key elements of the vertebra including posterior elements and the 3D-3D registration technique used for 3D-CT model can be applied (Fig.1). This type of methodology can be used in the clinic to evaluate with sufficient accuracy subject-specific spinal kinematics without exposure to additional radiation. The MRI-based 3D bone-marrow model may also be useful for kinematic analyses of other major joints such as hip, knee, ankle and shoulder joints.
