[Validation of the DTP-3 system for noninvasive spinal shape measurement by comparison with X-ray examination]

Repeated measurements of the spine are absolutely necessary in children and adolescents affected by spinal deformities especially during their growing-up periods. To avoid risks of tissue damage from x-ray exposure, several methods for non-invasive measurement of the spinal curvature have been developed. One of them is the DTP-3 position system allowing for a three-dimensional measurement of anatomical landmarks (spinous processes) and the calculation of curvature angles in both the frontal and sagittal planes. We were interested to know whether the DTP-3 was precise enough to determine the true spinal curvature.

To determine the precision of the DTP-3 system, we constructed a model of the spine. The model was then repeatedly investigated by both the noninvasive and x-ray methods. The distortion of x-ray images caused by the central projection mechanism was considered and included in the calculation. In addition, a group of patients with scoliosis up to 40° was evaluated by both the DTP-3 system and x-ray (the latter according to Cobb's method).

Differences in spatial coordinates between DTP-3 and x-ray examinations reached 20.9 mm in the frontal plane and 67.3 mm in the sagittal plane without distortion correction of x-ray images. The differences decreased below 1.5 mm after image distortion correction in each plane. Distortion correction had not the same effect for angle parameters as for coordinates. Differences between the DTP-3 angle parameters and Cobb's x-ray angles were below 4.7°, both without correction and after correction. The difference between DTP-3 angle parameters and Cobb's x-ray angles was -1.8° ± 3.0° (mean ± standard deviation) when measurement was performed on the patients with scoliosis.

The goal of any clinical examination is to obtain data applicable to decision-making analysis. In the case of scoliosis it is necessary to report results in terms of Cobb's angle, which is the problem for all surface-dependent methods, especially in patients with double curves. A solution may be to define the maximal difference between noninvasive and x-ray methods that could be acceptable for good clinical practice.

In this study we report good concordance between noninvasive and x-ray examinations of a modeled spinal deformity in terms of both angle and linear measurements. The same results were obtained for angle measurements in a group of patients with scoliosis up to 40°. Based on this study and our previous data we believe that the DTP-3 system can be introduced into clinical practice.