Introduction: Within the field of spinal fusion surgery, it is recognized that sub-optimal spinal alignment can accelerate degeneration of the spine, change of shape of the spine over time, and may increase the likelihood of revision surgery for the patient. 

The current standard of care to assess intraoperative spinal alignment is to either visually inspect the anatomy or use fluoroscopic imaging. There is currently no intraoperative solution that enables surgeons to precisely visualize changes in alignment without the use of potentially harmful ionizing radiation.

A novel real-time information software called Flash™ Align has been developed using machine-vision technology in conjunction with a registration algorithm requiring only visible light that matches the patient’s 3D intraoperative position with a preoperative CT scan to quantify intraoperative spinal alignment without radiation in the operating room. 

Methods: The results presented have been obtained via a single arm, single-center pilot study to test the Flash Align Software for the quantification of spinal alignment during posterior spine surgery. Three three-dimensional (3D) images were taken intraoperatively using the FLASH Navigation System, at three time points: (A) after patient positioning, (B) after decompression and interbody placement, and (C) after final positioning of the screws and interconnecting rods. These images were postoperatively analyzed using the Flash Align software where endplates were defined in 3D, a registration was simulated, and spinal parameters were calculated. The spinal alignment measurements from the Flash Align Software were compared to those of the participant’s intraoperative fluoroscopy images at the same three time points. Flash Align measurements from time (C) were also compared to a postoperative CT scan.

Results: In total, 16 patients were included in this study, however it was not possible to analyze all patients at each time point with all imaging modalities. The difference between the mean intraoperative sagittal Cobb angle using Flash Align and the fluoroscopy image at time (A) was 4.2 degrees (N=15, p=0.93); time (B) was 5.3 degrees (N=10, p=0.24); and time (C) was 5.2 degrees (N=10, p=0.33). 

A total of 10 patients were analyzed at time (C) comparing the calculated angle from Flash Align and the measured angle based on the postoperative CT scan. Four comparative measures were taken with absolute differences of; 1.7 degrees (p=0.80) in the sagittal Cobb angle, 2.2 degrees (p=0.38) in the coronal Cobb angle, 2.2 mm (p<0.5) in the coronal shift and 2.0 degrees (p=0.80) in the relative rotation.

Discussion: When comparing the sagittal and coronal Cobb angles calculated by Flash Align to those of fluoroscopic and CT imaging, the angles calculated using CT imaging modality presented a smaller mean difference. Two-dimensional fluoroscopic images lack critical information especially considering spinal deformity is characterized as a three-dimensional problem defined by sagittal and coronal alignment as well as axial rotation. By analyzing the spine in three dimensions, the endplate definition technique as employed in Flash Align is analogous to a standard post-operative CT scan analysis. Flash Align also enables a user to re-define the sagittal plane using anatomical landmarks such as aligning the femoral heads.