This study presents the next step from the study of  into developing an intra operative tool to monitor changes in external trunk geometry in real time during scoliosis surgery. This technique is appealing in several ways: 1) it provides 3-D data that can be correlated with the external shape of the patient during initial positioning on the operative table 2) it paves the way to provide continuous spatial tracking during the surgery without any interference or added radiation exposure to the patient, 3) it provides a simple and feasible workflow to be used for intraoperative setup, 4) It can provide direct real time feedback to the surgeon in order to evaluate the efficacy of his correction maneuvers, and 5) It could be used to compare the efficiency of different instrumentation systems or surgical maneuvers to correct scoliosis. In particular, in patients with significant global imbalance, pelvic or shoulder obliquity, the current technique could allow the surgeon to better assess these parameters during his correction maneuvers, so to decrease the use of intra-operative radiographs.
The optical sensors recorded only slight variations of rotations angles (5–10 degrees) on the Relton-Hall frame, where any displacement of the supports during the procedure is prohibited. It is known that about 30% of the correction is attributed by the positioning of the patient on the frame [14, 15]. Accordingly, continuous tracking of the trunk geometry during the positioning step is advantageous since it would allow optimal positioning and curve correction before the surgery. Moreover, it can provide insight about the patient's external trunk geometry under draping sheets, which cannot be assessed from radiographs. This becomes extremely useful to provide quantitative information about the positioning of the patient on the Relton-Hall frame before, during and after surgery. Therefore, the technique described above could assist the surgeon in obtaining a global view of the patient's deformation over the surgical field during positioning stage and all along the surgery. Although only AIS patients were evaluated in this study, the technique could be applied to any patient with spinal deformity, such as patients with neuromuscular scoliosis who often present with severe imbalance and pelvic obliquity.
This study presents an accurate intra-operative system that could be used to assist spine surgeons to evaluate patient positioning and external trunk correction. The workflow described in this study involves only minor modifications to the current surgical setup. Moreover, this technique is not harmful to the patient and no radiations are involved for providing intraoperative guidance. This is appealing to obtain real-time continuous data in an intra operative setting. Improvements from the previous study involving magnetic tracking  are: 1) the optoelectronic camera system does not interfere in any way with the current surgical setup. 2) the tracking system does not require any wiring to connect the markers, since passive spheres do not require any electrical input to be tracked in 3-D. 3) the optoelectronic camera are not subject to inaccuracy due to ferromagnetic material, hence continuous data acquisition is feasible during the surgery.
Trunk geometry can sometimes be difficult to assess due to displacement and elasticity of the skin. Many authors, mainly targeted for gait analysis  and for spinal motion measurement , proposed external landmark tracking for real-time assessment of the trunk. In these cases, skin displacement might occur when the motion is relatively large, which was not the case in our study. Displacement on the skin could potentially alter the true location of a landmark used to compute an angle. However, in the current study, the landmarks that are used to compute angles are all distant to each other and an error of even 1 cm will affect the resulting angle by less than 2 degrees if the landmarks are 30 cm apart. Also, by considering only indices after exposure of the spine, prior to instrumentation, it is possible to reduce the errors due to skin displacement and to obtain a more robust trunk tracking. Moreover, the patient is sedated and supported by the surgical frame. Only the corrective motion by the clinician and the respiratory function might alter the positioning of the landmarks on the patient. To verify this hypothesis, a larger number of landmarks, could be disposed in a uniform distribution on the patient, to quantify skin displacement local to each landmarks.