Do intraoperative radiographs in scoliosis surgery reflect radiographic result?

Posterior spinal fusion with lowest instrumented vertebra at L4 in idiopathic scoliosis: optimizing radiographic outcomes using pre-operative flexibility radiographs

PURPOSE

In idiopathic scoliosis (IS), there is general agreement ending PSFs at L3 or more cranial is preferred to optimize spinal motion, and extending PSFs to L4 may be necessary; however, this may also cause coronal imbalance or caudal disc wedging post-operatively due to leveling of L4 tilt. The purpose of this study was to identify a pre-operative radiographic measurement, which can be used to quantify the optimal amount of L4 tilt for ideal post-operative radiographic alignment.

METHODS

The study was a retrospective analysis of IS patients who underwent PSF to L4, with minimum 2-year follow-up post-operatively. Optimal outcome was defined by coronal balance, and L4-5 and L5-S1 disc wedging.

RESULTS

44 patients (84% females, mean age 13.6 years) were included. Analysis of pre-operative flexibility radiographs determined only the L5 tilt on the right side-bending (RSB) radiograph correlated with optimal outcome 2 (p = 0.03). To confirm the validity, the RSB value was subtracted from the post-operative C7-L4 tilt and the odds ratio analysis which was significantly correlated with optimal outcome 1 at final follow-up (OR 1.04, 95% CI 1-1.09).

CONCLUSIONS

In PSF to L4 for IS, L5 tilt measured from the pre-operative supine RSB radiograph can be used to optimize radiographic outcomes. Matching the pre-operative L5 tilt on RSB radiograph by leaving L4 tilted at the end of the PSF construct during surgery, quantified by the C7-L4 acute angle tilt, appears to be a useful method to achieve the desired post-operative alignment.

Multi-perspective region-based CNNs for vertebrae labeling in intraoperative long-length images

PURPOSE

Effective aggregation of intraoperative x-ray images that capture the patient anatomy from multiple view-angles has the potential to enable and improve automated image analysis that can be readily performed during surgery. We present multi-perspective region-based neural networks that leverage knowledge of the imaging geometry for automatic vertebrae labeling in Long-Film images - a novel tomographic imaging modality with an extended field-of-view for spine imaging.

METHOD

A multi-perspective network architecture was designed to exploit small view-angle disparities produced by a multi-slot collimator and consolidate information from overlapping image regions. A second network incorporates large view-angle disparities to jointly perform labeling on images from multiple views (viz., AP and lateral). A recurrent module incorporates contextual information and enforce anatomical order for the detected vertebrae. The three modules are combined to form the multi-view multi-slot (MVMS) network for labeling vertebrae using images from all available perspectives. The network was trained on images synthesized from 297 CT images and tested on 50 AP and 50 lateral Long-Film images acquired from 13 cadaveric specimens. Labeling performance of the multi-perspective networks was evaluated with respect to the number of vertebrae appearances and presence of surgical instrumentation.

RESULTS

The MVMS network achieved an F1 score of >96% and an average vertebral localization error of 3.3 mm, with 88.3% labeling accuracy on both AP and lateral images - (15.5% and 35.0% higher than conventional Faster R-CNN on AP and lateral views, respectively). Aggregation of multiple appearances of the same vertebra using the multi-slot network significantly improved the labeling accuracy (p < 0.05). Using the multi-view network, labeling accuracy on the more challenging lateral views was improved to the same level as that of the AP views. The approach demonstrated robustness to the presence of surgical instrumentation, commonly encountered in intraoperative images, and achieved comparable performance in images with and without instrumentation (88.9% vs. 91.2% labeling accuracy).

CONCLUSION

The MVMS network demonstrated effective multi-perspective aggregation, providing means for accurate, automated vertebrae labeling during spine surgery. The algorithms may be generalized to other imaging tasks and modalities that involve multiple views with view-angle disparities (e.g., bi-plane radiography). Predicted labels can help avoid adverse events during surgery (e.g., wrong-level surgery), establish correspondence with labels in preoperative modalities to facilitate image registration, and enable automated measurement of spinal alignment metrics for intraoperative assessment of spinal curvature.

Semiautomated intraoperative measurement of Cobb angle and coronal C7 plumb line using deep learning and computer vision for scoliosis correction: a feasibility study

OBJECTIVE

Scoliosis is a degenerative disease with a 3D deformity in the alignment of the spinal column. Surgical spinal correction outcomes are heavily dependent on the surgeon's expertise and use of visual cues because of time requirements, lack of automation, and radiation exposure associated with current intraoperative measurement techniques. In this study, the authors sought to validate a novel, nonradiographic, and semiautomated device that measures spinal alignment intraoperatively using deep learning and computer vision.

METHODS

To obtain spinal alignment metrics intraoperatively, the surgeon placed 3D-printed markers made of acrylonitrile butadiene styrene (ABS) plastic at designated locations in the surgical field. With the high-definition camera of the device, the surgeon can take an image of the markers in the surgical field. Images are processed through a computer vision model that detects the location of the markers and calculates the Cobb angle and coronal plumb line. The marker detection model was trained on 100 images and tested on 130 images of the ABS markers in various conditions. To verify the Cobb angle calculation, 50 models of angle templates from 0° to 180° in 3.6735° increments were created for testing. To verify the plumb line calculation, 21 models of plumb line measurements from -10 to +10 cm in increments of 1 cm were created for testing. A validation study was performed on a scoliotic cadaver model, and the radiographic calculations for Cobb angle and plumb line were compared with the device's calculations.

RESULTS

The area under the curve for the marker detection model was 0.979 for Cobb angle white, 0.791 for Cobb angle black, and 1 for the plumb line model. The average absolute difference between expected and measured Cobb angles on the verification models was 1.726° ± 1.259°, within the clinical acceptable error of 5°. The average absolute difference between the expected and measured plumb lines on the verification models was 0.415 ± 0.255 cm. For the cadaver validation study, the differences between the radiographic and device calculations for the Cobb angle and plumb line were 2.78° and 0.29 cm, respectively.

CONCLUSIONS

The authors developed and validated a nonradiographic, semiautomated device that utilizes deep learning and computer vision to measure spinal metrics intraoperatively.

Intraoperative stitched fluoroscopic images: effect of parallax on angular measurements of the spine

BACKGROUND CONTEXT

Intraoperative stitched O-arm images are commonplace during spinal deformity correction surgeries; however, the accuracy of stitched images for measuring angular measures is unknown.

PURPOSE

To examine the effect of radiographic parallax effect of stitched O-arm images by assessing the regional curve agreement with measurements from computed tomography (CT).

STUDY DESIGN/SETTING

Experimental radiographic study.

PATIENT SAMPLE

Four whole body cadavers (age: 81±14, sex: 2M/2F) and two fabricated spine model phantoms from surgical cases, one with extreme scoliosis and one normal spine, were utilized.

OUTCOME MEASURES

The limits of agreement for angular measures between CT (gold-standard) and intraoperative stitched fluoroscopic images were calculated. Further, intra- and inter-rater reliability was measured.

METHODS

A series of adjacent anterior-posterior and lateral images were acquired cranial to caudal using an O-arm in three table configurations (standard position, off-axis in the coronal plane, and reverse Trendelenburg) and stitched manually. Regional angular measures were extracted, and the limits of agreement were calculated between each table position and CT using a Bland-Altman approach.

RESULTS

The observers displayed excellent inter-rater reliability across table positions (range: 0.944-0.989) and intra-rater reliability (0.979-0.995). The limits of agreement results showed a similar and better agreement was observed for the Standard and Reverse Trendelenburg than the Off-Axis position.

CONCLUSIONS

This work shows reliable regional curvature measurements can be calculated with good agreement with CT in common table positions, but care should be taken to ensure the patient is perpendicular to the X-rays, particularly in the lateral view.

Can Postoperative Distal Adding-On be Predicted in Lenke Type 1B and 1C Curves with Intraoperative Radiographs?

STUDY DESIGN

A retrospective study of consecutive collected data.

OBJECTIVE

To investigate risk factors for postoperative distal adding-on (DA) in Lenke Type 1B and 1C curves using intraoperative radiographs.

SUMMARY OF BACKGROUND DATA

In adolescent idiopathic scoliosis (AIS), DA radiographic complication can negatively affect postoperative clinical results. However, few studies have focused on assessing risk factors for DA using intraoperative radiographs.

METHODS

We retrospectively evaluated 69 AIS patients with Lenke Type 1B or 1C curves who underwent posterior selective thoracic fusion. We divided patients into DA and non-DA groups based on radiograph data at 2-year follow-up using Wang et al (Spine 2011) definition of DA. We compared coronal radiographic parameters, including relative positions of end vertebra (EV), stable vertebra (SV), neutral vertebra (NV), and last touching vertebra (LTV) to lowest instrumented vertebra (LIV), and intraoperative radiographic parameters, between the two groups.

RESULTS

DA was present in 13 patients (18.8%) at 2-year follow-up. The mean LIV-EV, LIV-NV, LIV-SV, and LIV-LTV relative positions were significantly smaller in the DA group than in the non-DA group. Multivariate analysis showed that LIV-LTV was significantly associated with DA (DA: -0.2 ± 0.7, non-DA: 0.6 ± 0.7). Intraoperative radiographs showed that the mean angulation of the first disc below the LIV after final adjustment was significantly larger in the DA group (2.3° ± 1.1°) than in the non-DA group (0.9° ± 0.7°). Patients whose angulation of the first disc below the LIV was more than 3° were significantly associated with DA.

CONCLUSION

The LIV selected at more cranial to the LTV may be a risk factor for postoperative DA in Lenke Type 1B and 1C curves. Moreover, it was suggested that LIV extension might be considered when the first disc's angulation below the LIV is >3° in intraoperative radiographs.Level of Evidence: 3.

Prediction of long-term postoperative results of disc wedge and vertebral tilt with intraoperative prone radiograph in posterior correction of thoracolumbar/lumbar curve in adolescent idiopathic scoliosis: a minimum 5-year follow-up

BACKGROUND CONTEXT

Preservation of the more mobile lumbar segments is important during thoracolumbar/lumbar scoliosis surgery; however, the remaining disc wedge angle (DWA) below lowermost instrumented vertebra (LIV) and vertebral body tilt below LIV (LIV+1 tilt) can cause curve progression.

PURPOSE

This study aimed to evaluate the efficacy of intraoperative radiograph to predict the postoperative DWA below LIV and LIV+1 tilt on standing radiographs in patients with LIV of L3 or L4.

STUDY DESIGN/SETTING

Retrospective cohort study PATIENT SAMPLE: A total of 235 patients with idiopathic scoliosis who underwent posterior correction and fusion for the structural thoracolumbar curve and were followed up for >5 years were reviewed.

OUTCOME MEASURES

DWA below LIV, LIV+1 tilt, Cobb angle, trunk shift, apical vertebra translation, and pelvic parameters were measured.

METHODS

Correlation between intraoperative and postoperative measurements of DWA below LIV and LIV+1 tilt were assessed. Additional analysis was performed to identify risk factors and prognosis of LIV+1 tilt ≥10° and DWA below LIV of ≥4° RESULTS: LIV+1 tilt measured on intraoperative radiograph was significantly correlated with the postoperative 5-day and postoperative 5-year evaluation in both groups. However, the intraoperative DWA below LIV was only correlated with the postoperative 5-year value in the L3 group (p=.018). At the 5-year follow-up, patients with LIV+1 tilt ≥10° on intraoperative radiography showed significantly greater LIV+1 tilt (p<.001), apical vertebral translation (p<.001), thoracic curve (p=.008), and thoracolumbar curve (p<.001) than patients with LIV+1 tilt <10°. Intraoperative DWA below LIV of ≥4° was only associated with higher DWA below LIV at the 5-year follow-up.

CONCLUSIONS

Intraoperative measurement of LIV+1 tilt was correlated with long-term postoperative outcomes, and intraoperative LIV+1 tilt ≥10° was associated with a less favorable radiographic outcome. Intraoperative DWA below LIV demonstrated less correlation with postoperative values and was not a prognostic factor for other radiographic parameters.

Intraoperative versus postoperative radiographic coronal balance for adult spinal deformity surgery

BACKGROUND

Coronal malalignment in adult spinal deformity (ASD) has a close relationship with patient clinical outcomes. The purpose of this study is to evaluate the relationship between intra- and postoperative coronal radiographic parameters. A novel parameter, the central sacral pelvic line (CSPL), and its relation to the central sacral vertical line (CSVL) is explored. CSPL is a measure of spinal alignment referenced to the patient's pelvis as an intraoperative proxy for CSVL. CSVL is difficult to measure intraoperatively, because a C7-plumb line (referenced to gravity) cannot be drawn in the supine position.

METHODS

47 subjects ≥ 18 years old undergoing a spinal fusion of ≥ 6 levels from 2015 to 2017 were enrolled. The CSPL is defined as the perpendicular line bisecting the midpoint of the line that connects the superior aspects of the acetabuli. Two metrics describing coronal alignment were derived from each radiograph: (1) horizontal distance between the C7-plumb line and the CSPL at C7 (C7-CSPL) and (2) horizontal distance between the C7-plumb line and CSVL (C7-CSVL). Pearson's correlation and linear regression analysis was used to study the relationship between the intraoperative C7-CSPL and the postoperative C7-CSVL.

RESULTS

On average, the intraoperative C7-CSPL distance was 32.1 mm, postoperative C7-CSPL 20.8 mm, and postoperative C7-CSVL 18.9 mm. 15/47 (32%) had intraoperative C7-CSPL measurements > 4 cm, requiring intraoperative correction. Of those 15, 10 patients (67%) still had a postoperative C7-CSVL < 4 cm. Linear regression modeling indicates that when intraoperative CSPL is < 7.7 cm on average, the postoperative C7-CSVL will < 4 cm-our threshold for adequate coronal alignment. Patients with intraoperative C7-CSPL > 5 cm had a 50% chance of having a postoperative C7-CSVL > 4 cm; patients with intraoperative C7-CSPL < 5 cm had a 3% chance of having coronal malalignment. There is a strong positive relationship between postoperative C7-CSPL and C7-CSVL (r = 0.80 and 0.85, respectively).

CONCLUSION

In adult spinal surgery, the intraoperative coronal alignment measured using the novel C7-CSPL distance correlates well with postoperative C7-CSVL distance. This gives the surgeon an objective measurement of the correction they need after assessing initial intraoperative imaging. Our findings suggest an intraoperative C7-CSPL distance < 5 cm as a threshold value to predict postoperative C7-CSVL < 4 cm in 97% of patients tested.

Intraoperative Prone Radiographs During Scheuermann Kyphosis Correction Closely Estimate Standing Thoracic and Lumbar Parameters at 2 Years

BACKGROUND

In patients with Scheuermann kyphosis (SK) undergoing posterior spinal fusion with instrumentation (PSFI), intraoperative lateral radiographs assess deformity correction in the prone position. The relationship between thoracic and (partially un-instrumented) lumbar parameters on prone intraoperative versus standing postoperative radiographs is unknown.

METHODS

Forty-five consecutive patients with SK who underwent PSFI between 2007 and 2014 were reviewed. Thoracic kyphosis (TK), lumbar lordosis (LL), instrumented level kyphosis [upper instrumented vertebrae (UIV)-lower instrumented vertebrae (LIV)], and traditional sagittal parameters were recorded from preoperative standing, intraoperative prone, first outpatient standing, and >2-year standing radiographs and time periods were compared. Exclusion criteria included reduction modification after intraoperative radiographs and postoperative construct revision prohibiting comparison to initial intraoperative radiographs.

RESULTS

Twenty-five patients averaging 16 (12 to 20) years old during surgery with 3.1 (2 to 7) years follow-up met inclusion criteria. Average surgical variables included: 13±1 fusion levels, UIV at T2, LIV at L3, 3.8±1.6 osteotomies per patient, and 43±9% correction of TK. Preoperative TK and LL measured 82 and 76 degrees, respectively. TK on intraoperative (47 degrees), 6-week (49 degrees), and >2-year (50 degrees) radiographs changed significantly only between intraoperative and >2-year radiographs (P=0.03) by just 3 degrees. LL increased 5 degrees from intraoperative prone to 6-week standing radiographs (51 to 56 degrees, P=0.01) without further significant change at >2 years (59 degrees, P=0.09). Instrumented levels (UIV-LIV) had increased kyphosis at 6 weeks (32 to 35 degrees, P=0.01) without further change at >2 years (36 degrees, P=0.06).

CONCLUSIONS

TK on intraoperative prone radiographs during PSFI for SK should match the standing TK ∼6 weeks later. Intraoperative prone LL only slightly increases on early standing radiographs. Assuming a routine postoperative course, intraoperative radiographs slightly underestimate TK (by 3 degrees) and LL (by 8 degrees) on >2-year standing radiographs. These parameters (TK, LL, UIV-LIV) are visualized during surgery and should be used in future studies to predict long-term outcomes.

LEVEL OF EVIDENCE

Level IV-retrospective study.

Intraoperative image guidance for the surgical treatment of adult spinal deformity

Operative management of adult spinal deformity (ASD) has been increasing in recent years secondary to an aging society. The advance of intraoperative image guidance, such as the development of navigation and robotics systems has contributed to the growth and safety of ASD surgery. Currently, intraoperative image guidance is mainly used for pedicle screw placement and the evaluation of alignment correction in ASD surgery. Though it is expected that the use of navigation and robotics would result in increasing pedicle screw accuracy as reported in other spine surgeries, there are no well-powered studies specifically focusing on ASD surgery. Currently, deformity correction relies heavily on preoperative planning, however, a few studies have shown the possibility that intraoperative image modalities may accurately predict postoperative spinopelvic parameters. Future developments of intraoperative image guidance are needed to overcome the remaining challenges in ASD surgery such as radiation exposure to patient and surgeon. More novel imaging modalities may result in evolution in ASD surgery. Overall there is a paucity of literature focusing on intraoperative image guidance in ASD surgery, therefore, further studies are warranted to assess the efficacy of intraoperative image guidance in ASD surgery. This narrative review sought to provide the current role and future perspectives of intraoperative image guidance focusing on ASD surgery.

Is it mandatory to routinely use image intensifier during scoliosis surgery? – Results of an email survey

Background

Methods

Results

Conclusions

Automated Measurement of Lumbar Lordosis on Radiographs Using Machine Learning and Computer Vision

STUDY DESIGN

Cross sectional database study.

OBJECTIVE

To develop a fully automated artificial intelligence and computer vision pipeline for assisted evaluation of lumbar lordosis.

METHODS

Lateral lumbar radiographs were used to develop a segmentation neural network (n = 629). After synthetic augmentation, 70% of these radiographs were used for network training, while the remaining 30% were used for hyperparameter optimization. A computer vision algorithm was deployed on the segmented radiographs to calculate lumbar lordosis angles. A test set of radiographs was used to evaluate the validity of the entire pipeline (n = 151).

RESULTS

The U-Net segmentation achieved a test dataset dice score of 0.821, an area under the receiver operating curve of 0.914, and an accuracy of 0.862. The computer vision algorithm identified the L1 and S1 vertebrae on 84.1% of the test set with an average speed of 0.14 seconds/radiograph. From the 151 test set radiographs, 50 were randomly chosen for surgeon measurement. When compared with those measurements, our algorithm achieved a mean absolute error of 8.055° and a median absolute error of 6.965° (not statistically significant, P > .05).

CONCLUSION

This study is the first to use artificial intelligence and computer vision in a combined pipeline to rapidly measure a sagittal spinopelvic parameter without prior manual surgeon input. The pipeline measures angles with no statistically significant differences from manual measurements by surgeons. This pipeline offers clinical utility in an assistive capacity, and future work should focus on improving segmentation network performance.

Defining criteria for optimal lumbar curve correction following the selective thoracic fusion surgery in Lenke 1 adolescent idiopathic scoliosis: developing a decision tree

OBJECTIVE

The aim of this study was to identify the range of optimal versus suboptimal rates of spontaneous lumbar Cobb correction (SLCC%) and the factors predicting such outcomes in a cohort of Lenke 1 adolescent idiopathic scoliosis (AIS) after posterior spinal fusion surgery.

METHODS

Seventy-one consecutive Lenke1 B and C AIS patients with a fusion level to L1 and higher with two-year follow-up were included. Thoracic kyphosis (T1-T4 and T4-T12 TK), lumbar lordosis (L1-S1 LL), thoracic and lumbar Cobb angles, thoracic and lumbar apical vertebral rotations and translations (AVR and AVT), pelvic incidence, sacral slope, and sagittal and frontal balances were measured at preoperative, early postoperative, and two-year follow-up. The SLCC% was calculated between preoperative and two-year follow-up. A clustering analysis determined the subgroups of patients with significantly higher and lower (optimal versus suboptimal) rate of SLCC% in the cohort at two-year follow-up. The cutoff values of the preoperative and early postoperative radiographic parameters that significantly predicted the optimal and suboptimal SLCC% were determined using a decision tree.

RESULTS

The averages of the optimal versus suboptimal range of SLCC% in the cohort were 72% [55%, 105%] versus 39% [- 7%, 42%]. Preoperative and early postoperative spinal parameters predicted the optimal versus suboptimal SLCC% with an accuracy of 82%, 95%CI [0.73-0.94]. Preoperative AVT_Lumbar < 10 mm was a predictor of optimal SLCC%. In patients with a preoperative AVT_Lumbar > 10 mm, early postoperative T4-T12 TK < 24° (but not less than 17°) accompanied by - 5° < AVR_Thoracic < 5° were the main predictors of optimal SLCC% in our cohort.

CONCLUSION

Quantitative clustering of the SLCC% into optimal and suboptimal groups allowed identifying the cutoff values of preoperative (AVT_Lumbar) and early postoperative (T4-T12 TK and AVR_Thoracic) spinal parameters that can predict the optimal range of SLCC% at two-year postoperative in our cohort of Lenke 1 AIS.

LEVEL OF EVIDENCE

IV.

Measurement of Spinopelvic Angles on Prone Intraoperative Long-Cassette Lateral Radiographs Predicts Postoperative Standing Global Alignment in Adult Spinal Deformity Surgery

STUDY DESIGN

Retrospective review from a single institution.

OBJECTIVES

To evaluate intraoperative T1-pelvic angle (TPA), T4PA, and T9PA as predictors of postoperative global alignment after adult spinal deformity (ASD) surgery.

SUMMARY OF BACKGROUND DATA

Malalignment following adult spinal surgery is associated with disability and correlates with health-related quality of life. Preoperative planning and intraoperative verification are crucial for optimal postoperative outcomes. Currently, only pelvic incidence minus lumbar lordosis (PI-LL) mismatch has been used to assess intraoperative correction.

METHODS

Patients undergoing ≥4-level spinal fusion with full-length pre-, intra-, and first postoperative calibrated radiographs were included from a single institution. Alignment measurements were obtained for sagittal vertical axis (SVA), PI-LL, TPA, T4PA, and T9PA. The whole cohort was divided into upper thoracic (UT: UIV > T7) and lower thoracic fusions (LT: UIV < T7). Change was assessed between phases, and a subanalysis was included for UT and LT groups to compare alignment changes for differing extent of proximal fusion in the sagittal plane.

RESULTS

Eighty patients (mean 63.4 years, 70% female, mean levels fused 11.9) underwent significant ASD correction (ΔPI-LL = 22.1°; ΔTPA = 13.8°). For all, intraoperative TPA, T4PA, and T9PA correlated with postoperative SVA (range, r = 0.41-0.59), whereas intraoperative PI-LL correlated less (r = 0.38). For UT (n = 49), all spinopelvic angles and LL were similar intraoperative to postoperatively (p > .09). For LT (n = 31), intraoperative and postoperative T9PA and LL were similar (p > .10) but TPA and T4PA differed (p < .02). For UT, all intraoperative and postoperative spinopelvic angles strongly correlated (r = 0.8-0.9). For LT, intraoperative to postoperative T9PA strongly correlated (r = 0.83) and TPA, T4PA, and LL correlated moderately (r = 0.65-0.70). LT trended toward more reciprocal kyphosis postoperatively (8.1° vs. 2.6°; p = .059).

CONCLUSIONS

Intraoperative measurements of TPA, T4PA, and T9PA correlated better with postoperative global alignment than PI-LL, demonstrating their utility in confirming alignment goals. When comparing intraoperative to postoperative films, only T9PA was similar in LT whereas all spinopelvic angles were similar in UT. Reciprocal kyphosis in unfused segments of LT fusions may account for difference in TPA and T4PA from intraoperative to postoperative films.

LEVEL OF EVIDENCE

Level III.

Unplanned Immediate Return to Operating Room After Spine Surgery: Significance of Immediate Postoperative Radiographs

STUDY DESIGN

Retrospective analysis; single center data.

OBJECTIVE

The purpose of this study is to look at the utility and relevance of immediate postoperative radiographs in providing vital information leading to immediate revision after spine surgery.

SUMMARY OF BACKGROUND DATA

Immediate postoperative radiographs are routinely obtained in the recovery room after spine surgery to verify the level, alignment of the spine, implant position, and the adequacy of the procedure. However, with the ability to utilize intraoperative fluoroscopy imaging for the same purpose, the requirement for immediate postoperative radiographs needs to be validated. The purpose of this study is to look at the utility and relevance of these postoperative radiographs in providing critical information that may warrant immediate intervention.

METHODS

Retrospective analysis of all spine surgeries (elective and emergent), performed at a single center from 2011 to 2016, was done and cases returning to operating room within 48 hours were identified. Indication of immediate revision was reviewed and utility of immediate postoperative radiographs in guiding immediate revision was analyzed.

RESULTS

A total of 1804 elective and urgent spinal surgeries were performed by seven surgeons. Twenty-two patients returned to operating room within 48 hours of their index procedures. Of these 22 cases, only two patients were noted to have positive findings on recovery room radiographs. The findings of suboptimal spinal alignment or failed instrumentation led to the immediate revision in both cases. Both cases involved instrumentation at cervicothoracic region and intraoperative imaging provided only limited visualization.

CONCLUSION

Routine recovery room radiographs played a role in the decision to emergently return to the operating room in 0.10% (2/1804) cases at our institution. The potential benefit of immediate recovery room radiographs after spine surgery should be weighed against the added healthcare cost and patient discomfort associated with obtaining these radiographs routinely. Imaging may be delayed to a more elective time without any significant risk in majority of spine cases.

LEVEL OF EVIDENCE

3.

Assessment of Coronal Spinal Alignment for Adult Spine Deformity Cases After Intraoperative T Square Shaped Use

STUDY DESIGN

Retrospectively reviewed surgeries between 2011 and 2015 of patients who underwent posterior spinal deformity instrumentation with constructs involving fusions to pelvis and encompassing at least five levels.

OBJECTIVE

Measure the radiographic outcomes of coronal malalignment (CM) after use of an intraoperative T square shaped instrument in posterior spinal deformity surgeries with at least five levels of fusion and extension to pelvis.

BACKGROUND

Neuromuscular children found to benefit from intraoperative T square technique to help achieve proper coronal spinal balance with extensive fusions. This intraoperative technique used in our posterior spine deformity instrumentation surgeries with the aforementioned parameters.

METHODS

There were 50 patients: n = 16 with intraoperative T square and n = 34 no-T square shaped device. Subgroups divided based on greater than 20 mm displacement and greater than 40 mm displacement of the C7 plumb line to the central sacral vertical line on either side in preoperative radiographs. We analyzed the demographics and the pre- and postoperative radiographic parameters of standing films: standing CM (displacement of C7 plumb line to central sacral vertical line), and major coronal Cobb angles in total sample and subgroups and compared T square shaped device with no-T square shaped device use by analysis of variance. A p value ≤.05 is statistically significant.

RESULTS

In the total sample, though postoperative CM mean was not statistically different, we observed greater CM corrections in patients where a T square shaped device was used (70%) versus no-T square shaped device used (18%). In >20 mm and >40 mm subgroups, the postoperative mean CM values were statistically lower for the patients where a T square shaped device was used, p = .016 and p = .003, respectively. Cobb corrections were statistically higher for T square shaped device use in both >20 mm and >40 mm subgroups, 68%, respectively.

CONCLUSION

The intraoperative T square shaped device technique had a positive effect on the amount of spine coronal malalignment correction after its use and for lumbar and thoracic coronal Cobb angles.

LEVEL OF EVIDENCE

Level III.

Data-driven Classification of the 3D Spinal Curve in Adolescent Idiopathic Scoliosis with an Applications in Surgical Outcome Prediction

Adolescent idiopathic scoliosis (AIS) is a three-dimensional (3D) deformity of the spinal column. For progressive deformities in AIS, the spinal fusion surgery aims to correct and stabilize the deformity; however, common surgical planning approaches based on the 2D X-rays and subjective surgical decision-making have been challenged by poor clinical outcomes. As the suboptimal surgical outcomes can significantly impact the cost, risk of revision surgery, and long-term rehabilitation of adolescent patients, objective patient-specific models that predict the outcome of different treatment scenarios are in high demand. 3D classification of the spinal curvature and identifying the key surgical parameters influencing the outcomes are required for such models. Here, we show that K-means clustering of the isotropically scaled 3D spinal curves provides an effective, data-driven method for classification of patients. We further propose, and evaluate in 67 right thoracic AIS patients, that by knowing the patients’ pre-operative and early post-operation clusters and the vertebral levels which were instrumented during the surgery, the two-year outcome cluster can be determined. This framework, once applied to a larger heterogeneous patient dataset, can further isolate the key surgeon-modifiable parameters and eventually lead to a patient-specific predictive model based on a limited number of factors determinable prior to surgery.

Validation of prone intraoperative measurements of global spinal alignment

OBJECTIVE

Verifying the adequacy of surgical correction of adult spinal sagittal deformity (SSD) leads to improved postoperative alignment and clinical outcomes. Traditionally, surgeons relied on intraoperative measurements of lumbar lordosis (LL) correction. However, T-1 pelvic angle (TPA) and its component angles more reliably predict postoperative alignment. While TPA is readily measured on standing radiographs, intraoperative radiographs offer poor resolution of the bicoxofemoral axis. A method to recreate this radiographic landmark by extrapolating preoperative measurements has been described. The authors aimed to assess the reliability of measurements of global spinal alignment obtained via geometrical reconstitution of the bicoxofemoral axis on prone intraoperative radiographs.

METHODS

A retrospective review was performed. Twenty sets of preoperative standing full-length and intraoperative prone 36-inch lateral radiographs were analyzed. Pelvic incidence (PI) and sacral to bicoxofemoral axis distance (SBFD) were recorded on preoperative films. A perpendicular line was drawn on the intraoperative radiograph from the midpoint of the sacral endplate. This was used as one limb of the PI, and the second limb was digitally drawn at an angle that reproduced the preoperatively obtained PI, extending for a distance that matched the preoperative SBFD. This final point marked the obscured bicoxofemoral axis. These landmarks were used to measure the L-1, T-9, T-4, and T-1 pelvic angles (LPA, T9PA, T4PA, and TPA, respectively) and LL. Two spine fellows and 2 attending spine surgeons made independent measurements and repeated the process in 1 month. Mixed-model 2-way intraclass correlation coefficient (ICC) and Cronbach’s α values were calculated to assess interobserver, intraobserver, and scale reliability.

RESULTS

Interobserver reliability was excellent for preoperative PI and intraoperative LPA, T9PA, and T4PA (ICC = 0.88, 0.84, 0.84, and 0.93, respectively), good for intraoperative TPA (ICC = 0.68), and fair for preoperative SBFD (ICC = 0.60) and intraoperative LL (ICC = 0.50). Cronbach’s α was ≥ 0.80 for all measurements. Measuring PI on preoperative standing images had excellent intraobserver reliability for all raters (ICC = 0.89, range 0.80–0.93). All raters but one showed excellent reliability for measuring the SBFD. Reliability for measuring prone LL was good for all raters (ICC = 0.71, range 0.64–0.76). The LPA demonstrated good to excellent reliability for each rater (ICC = 0.76, range 0.65–0.81). The thoracic pelvic angles tended to be more reliable at more distal vertebrae (T9PA ICC = 0.71, range 0.49–0.81; T4PA ICC = 0.62, range 0.43–0.83; TPA ICC = 0.56, range 0.31–0.86).

CONCLUSIONS

Intraoperative assessment of global spinal alignment with TPA and component angles is more reliable than intraoperative measurements of LL. Reconstruction of preoperatively measured PI and SBFD on intraoperative radiographs effectively overcomes poor visualization of the bicoxofemoral axis. This method is easily adopted and produces accurate and reliable prone intraoperative measures of global spinal alignment.

Predicting massive transfusion in adolescent idiopathic scoliosis patients undergoing corrective surgery: Association of preoperative radiographic findings

Corrective surgery with a posterior approach for adolescent idiopathic scoliosis (AIS) is often accompanied by considerable bleeding. Massive transfusion after excessive hemorrhage is associated with complications such as hypothermia, coagulopathy, and acid-base imbalance. Therefore, prediction and prevention of massive transfusion are necessary to improve the clinical outcome of AIS patients. We aimed to identify the factors associated with massive transfusion in AIS patients undergoing corrective surgery. We also evaluated the clinical outcomes after massive transfusion.We included and analyzed AIS patients who underwent corrective surgery with a posterior approach from January 2008 to February 2015. We retrospectively reviewed the electronic medical records of 765 consecutive patients. We performed multivariable logistic regression analysis to assess the factors related to massive transfusion. Furthermore, we compared the effects of massive transfusion on clinical outcomes, including postoperative morbidity and hospital stay.Of 765 patients, 74 (9.7%) received massive transfusion. Body mass index (odds ratio [OR] 0.782, 95% confidence interval [CI] 0.691-0.885, P < .001) and the number of fused vertebrae (OR 1.322, 95% CI 1.027-1.703, P = .03) were associated with massive transfusion. In the comparison among the different Lenke curve types, Lenke type 4 showed the highest prevalence of massive transfusion. Patients in the massive transfusion group showed a higher incidence rate of postoperative morbidity and prolonged hospital stay.Massive transfusion was required in 9.7% of AIS patients who underwent corrective surgery with a posterior approach. A lower body mass index and higher number of fused vertebrae were associated with massive transfusion. Massive transfusion is related to poor clinical outcomes in AIS patients.

Less Invasive Surgical Correction of Adult Degenerative Scoliosis, Part I

BACKGROUND

Adult scoliosis is a condition with increasing prevalence and medical and socioeconomic importance. Surgery is fraught with a significant complication rate in an elderly multimorbid patient population.

OBJECTIVE

To assess technical feasibility and radiographic results of image-guided less invasive correction of adult degenerative scoliosis.

METHODS

Thirty individuals (age, 64–88 years) with progressive deformity (coronal Cobb angles &gt; 25° and &lt; 85°), intractable back pain, radiculopathy, or neurogenic claudication were treated by less invasive decompression and fusion (unilateral transforaminal interbody cage instrumentation and bilateral facet fusions) with recombinant human bone morphogenetic protein-2, spanning 3 to 8 segments (average, 6 segments), using biplanar fluoroscopy or intraoperative computed tomography (iCT)—based navigation. Accuracy of screw placement, curve correction, and fusion rate were evaluated during a mean follow-up of 19.6 months.

RESULTS

With 415 screws implanted, misplacement (grade II or greater) was not observed, and no implants required revision. Spinal iCT with automated registration required 17.5 ± 8.5 minutes (single registration for all segments); monosegmental bilateral screw insertion required 6.8 ± 3.4 minutes. Mean sagittal (coronal) Cobb angle correction was 44.8 ± 10.7° (31.7 ± 13.7°). Mean lumbar lordosis increased from 8.8 ± 8.9° to −36 ± 6.9°, and sagittal balance was reduced from 31.6 ± 15.2 to 8 ± 8.4 mm. Solid fusion was confirmed in 90% of instrumented segments at 16 months. Average radiation dose to the surgeon was 0.025 mSv for single-level transforaminal lumbar interbody fusion with fluoroscopic guidance vs 0 mSv with iCT navigation.

CONCLUSION

Instrumented correction of adult deformity was significantly facilitated by iCT navigation, eliminating radiation exposure to the surgeon. Intraoperative biplanar CT scout views including pelvis and shoulders allow comprehensive assessment of multiplanar deformity correction. Fusion rates obtained with less invasive access equal those of conventional open technique.