gms | German Medical Science

Objective: Spinal balance is a condition in which body weight is positioned along a line slightly behind the axis of rotation of the two femoral heads. This condition leads to an optimal dispension of weight and a state of minimal energy-consumption. Operations on the spine, especially instrumentations, can influence the spinal balance, and can have negative or positive long-term side effects on adjoining segments of the spine. For this reason planning of operations adjusted to the individual anatomy is necessary, especially in cases of compensation. Therefore, freehanded standing full spine x-rays are acquired and parameters are calculated. In order to speed up standard parameter calculation and to achieve a comprehensive set of lateral spinal parameters quickly, a semi-automated framework for spinal parameter calculation is presented.

Method: Within the developed framework for fast semi-automated spinal parameter calculation the user is enabled to manually place characteristic landmarks for each vertebra, defining cranial and caudal endplates. In case of lumbar or full spine radiographic data femoral heads are approximated using manually placed circles. In this process, each time new landmarks are placed, parameter calculation is updated by recalculation of all so far achievable parameters according to the set landmarks. Calculated spinal parameters include thereby standard parameters for sagittal balance like C7 plumb line, pelvic tilt/incidence, sacral slope, lumbar lordosis and thoracic kyphosis. Additionally, several vertebra specific parameters like vertebral wedge, cranial/caudal/mean depth and several segment specific parameters like midplane angle, disc height, sagittal plane displacement and Cobb’s angle are calculated automatically within. In this way more than 250 parameters are calculated for the whole spine. To achieve information about efficiency of the developed tool, 10 full spine lateral x-ray images were used.

Results: Placement of all landmarks (full spine) within the developed tool, including placement of femoral head approximation, and thereby calculation of a comprehensive set of spinal parameters took around 3.60±0.97 minutes. Manual measurement of a small subset of parameters (C7 plumb line, pelvic parameters, lordosis and kyphosis) already took around 10 minutes.

Conclusions: The presented semi-automated tool offers an efficient way to achieve a comprehensive set of spinal parameters that could be considered in planning of spinal surgery.