Automatic Lenke classification of adolescent idiopathic scoliosis with deep learning

Purpose

The Lenke classification system is widely utilized as the preoperative evaluation protocol for adolescent idiopathic scoliosis (AIS). However, manual measurement is susceptible to observer-induced variability, which consequently impacts the evaluation of progression. The goal of this investigation was to develop an automated Lenke classification system utilizing innovative deep learning algorithms.

Methods

Using the database from the First Affiliated Hospital of Sun Yat-sen University, the whole spinal x-rays images were retrospectively collected. Specifically, images collection was divided into AIS and control group. The control group consisted of individuals who underwent routine health checks and did not have scoliosis. Afterwards, relative features of all images were annotated. Deep learning was implemented through the utilization of the key-point based detection method to realize the vertebral detection, and Cobb angle measurement and scoliosis classification were performed based on relevant standards. Besides, the segmentation method was employed to achieve the recognition of lumbar vertebral pedicle to determine the type of lumbar spine modifier. Finally, the model performance was further quantitatively analyzed.

Results

In the study, a total of 2082 spinal x-ray images were collected from 407 AIS patients and 227 individuals in the control group. The model for vertebral detection achieved an F1-score of 0.809 for curve type evaluation and an F1-score of 0.901 for thoracic sagittal profile. The intraclass correlation efficient (ICC) of the Cobb angle measurement was 0.925. In the analysis of performance for vertebra pedicle segmentation model, the F1-score of lumbar modification profile was 0.942, the intersection over union (IOU) of the target pixels was 0.827, and the Hausdorff distance (HD) was 6.565 ± 2.583 mm. Specifically, the F1-score for ultimate Lenke type classifier was 0.885.

Conclusions

This study has constructed an automated Lenke classification system by employing the deep learning networks to achieve the recognition pattern and feature extraction. Our models require further validation in additional cases in the future.

Bracing in severe skeletally immature adolescent idiopathic scoliosis: does a holding strategy change the surgical plan?

PURPOSE

The purpose of the study was to assess the changes in flexibility during night-time bracing in skeletally immature adolescent idiopathic scoliosis (AIS) with curves in the surgical range.

MATERIALS AND METHODS

We included a consecutive cohort of 89 AIS patients with curves ≥ 45° and an estimated growth potential. All patients were eventually treated with fusion surgery, and all patients had side-bending radiographs prior to both bracing and surgery. Curves were classified as structural or non-structural curves according to Lenke at both timepoints.

RESULTS

The main curve progressed by a mean of 12 ± 10° and the secondary curve by 8 ± 8°. Flexibility of the main curve decreased from 50 ± 19% to 44 ± 19% (p = 0.001) and the underlying curve from 85 ± 21% to 77 ± 22% (p = 0.005). In 69 patients (79%), the Lenke category did not progress during bracing. In 14 patients (15%), the progression in Lenke type occurred in the thoracic region (i.e., Lenke type 1 to type 2), while six patients (7%) progressed in the lumbar region (i.e., type 1 to type 3). In the 69 patients that did not progress, we found that the last touched vertebra moved distally by one or two levels in 26 patients.

CONCLUSIONS

This is the first study to describe that curve flexibility decreases during bracing in severe AIS. However, this had only a modest impact on the surgical strategy. Bracing as a holding strategy can be applied, but the risk of losing flexibility in the lumbar spine should be outweighed against the risks of premature fusion surgery.

Anterior fusion surgery with overcorrection in the treatment of adolescent idiopathic scoliosis with Lenke 1 AR curve type: how to achieve overcorrection and its effect on postoperative spinal alignment

BACKGROUND

The efficacy of anterior fusion with overcorrection in the instrumented vertebra for Lenke 1 AR type curves has been reported, but how to achieve overcorrection and how overcorrection affects spinal alignment are unclear. The purpose of this study was to identify the factors that cause overcorrection, and to investigate how overcorrection affects postoperative spinal alignment in the surgical treatment of Lenke 1 AR type curves.

METHODS

Patients who had anterior surgery for a Lenke type 1 or 2 and lumbar modifier AR (L4 vertebral tilt to the right) type scoliosis and minimum 2-year follow-up were included. The radiographic data were measured at preoperative, postoperative 1 month, and final follow-up. The UIV-LIV Cobb angle was determined as the Cobb angle between the upper instrumented vertebra (UIV) and the lower instrumented vertebra (LIV), and a negative number for this angle was considered overcorrection. The screw angle was determined to be the sum of the angle formed by the screw axis and the lower and upper endplates in the LIV and UIV, respectively. The change (Δ) in the parameters from postoperative to final follow-up was calculated. The relationships between the UIV-LIV Cobb angle and other radiographic parameters were evaluated by linear regression analyses.

RESULTS

Fourteen patients met the inclusion criteria. Their median age was 15.5 years, and the median follow-up period was 53.6 months. The median UIV-LIV Cobb angle was -1.4° at postoperative 1 month. The median screw angle was 4.7°, and overcorrection was achieved in 11 (79%) cases at postoperative 1 month. The screw angle (r2 = 0.42, p = 0.012) and Δ FDUV-CSVL (the deviation of the first distal uninstrumented vertebra from the central sacral vertical line, r2 = 0.53, p = 0.003) were significantly correlated with the UIV-LIV Cobb angle.

CONCLUSIONS

Screw placement in the UIV and LIV not parallel to the endplate, but angled, was an effective method to facilitate overcorrection in the instrumented vertebrae. The results of the present study suggest that overcorrection could bring spontaneous improvement of coronal balance below the instrumented segment during the postoperative period.

Surgical Treatment of Adolescent Idiopathic Scoliosis with the ApiFix Minimal Invasive Dynamic Correction System-A Preliminary Report of a 24-Month Follow-Up

Adolescent idiopathic scoliosis (AIS) is a three-dimensional growth disorder. Corrective surgical procedures are the recommended treatment option for a thoracic angle exceeding 50° and a lumbar major curve of 40°. Over the past few years, dynamic growth modulation implants have been developed as alternatives to permanent fusion. The ApiFix system was designed as a 2D "posterior dynamic device" for curve correction. After implantation in a minimally invasive procedure, it uses polyaxial joints and a self-adjusting rod to preserve the degree of motion and to accommodate the patient's growth. It provides an effective method of controlling deformity and fills the gap between the conservative treatment of major curves that are >35° and the fusion procedure. The objective of the two-center cohort study was the analysis of the correction results of patients, who underwent surgical intervention with the ApiFix system. The inclusion criteria were AIS, Lenke type 1 or type 5, a major curve on bending films of ≤30°, and an angle of the major curve of between 35° and 60°. Postoperative radiograph data were obtained longitudinally for up to 24 months of follow-up and compared to preoperative (preop) values. For comparisons of the different time points, non-parametric tests (Wilcoxon) or paired t-tests for normally distributed values were used to analyze repeated measures. Overall, 36 patients (25 female and 11 male) were treated with the ApiFix system from April 2018 to October 2020. Lenke type 1 was identified in 21 (58%) cases and Lenke type 5 was identified in 15 (42%) cases. The average angle of the thoracic major curve for Lenke 1 was 43°. The preoperative lumbar major curve (Lenke 5) was determined to be 43°. Over a follow-up of 24 months, a correction of the major curve to an average of 20° was observed for Lenke 1 and that to an average of 15° was observed for Lenke 5. Lenke type 1 and type 5 showed significant changes in the major curve over the individual test intervals in the paired comparisons compared to the starting angle (Lenke 1: preop-24 months, 0.002; Lenke 5: preop-24 months, 0.043). Overall, 11 events were recorded in the follow-up period, that required revision surgery. We distinguished between repeated interventions required after reaching the maximum distraction length of the implant due to the continued growth of the patient (n = 4) and complications, such as infections or problems associated with the anchorage of the implant (n = 7). The results from the present cohort revealed a statistically significant improvement in the postoperatively measured angles of the major and minor curves in the follow-up after 24 months. Consequently, the results were comparable to those of the already established vertebral body tethering method. Alignment in AIS via dynamic correction systems in combination with a possible growth modulation has been a treatment alternative to surgical fusing procedures for more than a decade. However, the long-term corrective effect has to be validated in further studies.

Semi-automatic method for pre-surgery scoliosis classification on X-ray images using Bending Asymmetry Index

PURPOSE

Bending Asymmetry Index (BAI) has been proposed to characterize the types of scoliotic curve in three-dimensional ultrasound imaging. Scolioscan has demonstrated its validity and reliability in scoliosis assessment with manual assessment-based X-ray imaging. The objective of this study is to investigate the ultrasound-derived BAI method to X-ray imaging of scoliosis, with supplementary information provided for the pre-surgery planning.

METHODS

About 30 pre-surgery scoliosis subjects (9 males and 21 females; Cobb: 50.9 ± 19.7°, range 18°-115°) were investigated retrospectively. Each subject underwent three-posture X-ray scanning supine on a plain mattress on the same day. BAI is an indicator to distinguish structural or non-structural curves through the spine flexibility information obtained from lateral bending spinal profiles. BAI was calculated semi-automatically with manual annotation of vertebral centroids and pelvis level inclination adjustment. BAI classification was validated with the scoliotic curve type and traditional Lenke classification using side-bending Cobb angle measurement (S-Cobb).

RESULTS

82 curves from 30 pre-surgery scoliosis patients were included. The correlation coefficient was R² = 0.730 (p < 0.05) between BAI and S-Cobb. In terms of scoliotic curve type classification, all curves were correctly classified; out of 30 subjects, 1 case was confirmed as misclassified when applying to Lenke classification earlier, thus has been adjusted.

CONCLUSION

BAI method has demonstrated its inter-modality versatility in X-ray imaging application. The curve type classification and the pre-surgery Lenke classification both indicated promising performances upon the exploratory dataset. A fully-automated of BAI measurement is surely an interesting direction to continue our endeavor. Deep learning on the vertebral-level segmentation should be involved in further study.

Preoperative supine traction radiographs often result in higher Lenke classifications than supine bending radiographs in adolescent idiopathic scoliosis

STUDY DESIGN

Retrospective chart and radiographic review.

OBJECTIVE

The purpose of this study is to determine if both traction and side-bending radiographs yield the same Lenke classification. Supine side-bending radiographs are used to evaluate curve flexibility and assign Lenke classification in Adolescent Idiopathic Scoliosis (AIS). Supine traction radiographs are another tool used by treating surgeons to gauge flexibility and appropriate levels for spinal fusion in AIS.

METHODS

Retrospective chart and radiographic review were performed on AIS patients that underwent a posterior spinal fusion from 2008 to 2017. Cobb angles and Lenke classifications were determined on all upright posterioanterior (PA) spine radiographs, supine traction radiographs, and four supine bending radiographs. Statistical analysis using independent t tests and chi-square tests as appropriate were compared between patients with or without discordant Lenke classifications with p value set at < 0.05 for statistical significance.

RESULTS

184 patients met inclusion criteria, 36 males and 148 females. The average Cobb angle for the proximal thoracic (PT) curve was 27.2°, main thoracic (MT) curve was 60.5°, and thoracolumbar/lumbar (TL/L) curve was 48.0°. Significantly less curve correction was found with supine traction radiographs compared with bending radiographs: PT (23.1° vs 18.9°, p < 0.001), MT (38.9° vs 37.9°, p = 0.015), and TL/L (25.9° vs. 18.0°, p < 0.001). Lenke Classification was found concordant in 151/184 (82.1%). Traction views in the discordant Lenke classification group demonstrated less curve correction than those in the concordant group: PT (27.4° vs. 22.1°, p = 0.011), MT (45.3° vs. 37.5°, p < 0.001), and TL/L (29.3° vs 25.1°, p = 0.019).

CONCLUSION

Supine traction and supine bending radiographs provided a concordant Lenke classification 82.1% of the time. However, supine traction radiographs demonstrate less curve correction, a higher Lenke classification, and underestimated the TL/L curve correction to a greater degree. A single supine traction film is not an adequate substitute to side-bending radiographs when determining Lenke classification in patients with Adolescent Idiopathic Scoliosis.

LEVEL OF EVIDENCE

III.

Use of Temporary Rod and Intraoperative Neuroimaging for Correction of Adolescent Idiopathic Scoliosis: 2-Dimensional Operative Video

Adolescent idiopathic scoliosis (AIS) is an abnormal lateral curvature of the spine that arises during the pubescent growth spurt. AIS mainly affects females in the age group of 10 to 16 yr, with a prevalence of about 1% to 3% in the at-risk population.1 Treatment options vary depending on disease presentation and severity. Mild curvature mainly requires periodic observation for disease progression, whereas more moderate curvature can necessitate bracing or corrective surgery.2 Here, we present the use of a temporary rod and neuroimaging for the correction of Lenke type 1 spinal curvature in an AIS patient. An inferior facetectomy is performed, and a Lenke probe is used for entry into the pedicle and vertebral body. The channel is sounded before and after tapping to check for adequate walls prior to insertion of the pedicle screw. The ARTIS pheno is brought into the field, which uses fluoroscopy to create a 3-dimensional (3D) representation of the instrumentation within the spine. A temporary rod is placed in the concavity, and a combination of corrective techniques, including a rod roll, apical translation, and reduction, is performed to bring up the concavity, derotate the spine, and translate the spine. The permanent rod is then placed in the convexity, and compression is performed to bring down the scapular prominence. The temporary rod is then removed, and a final rod is placed to complete the 3D correction. The patient consented to the procedure, and an informed written consent was obtained from the patient to use her photographs and video recordings for publication.

Combination of Side-Bending and Traction Radiographs Do Not Influence Selection of Fusion Levels Compared to Either One Alone in Adolescent Idiopathic Scoliosis

STUDY DESIGN

A retrospective study.

OBJECTIVES

Curve flexibility in patients with adolescent idiopathic scoliosis (AIS) can be evaluated using different techniques. This study aimed to determine whether the combination of side-bending (SB) and traction (TX) radiographs influences preoperative planning for AIS than either radiograph alone.

METHODS

Thirty-two spine surgeons were asked to review 30 AIS Lenke type 1 cases and select an upper instrumented vertebra (UIV) and lower instrumented vertebra (LIV) for the posterior spinal instrumentation of each case. Each rater reviewed the cases 3 times in each round. The raters were provided with the full-length posteroanterior (PA) and lateral standing and SB radiographs for round 1; PA, lateral, and TX radiographs for round 2; and PA, lateral, SB, and TX radiographs for round 3. Intra- and inter-rater reliabilities were evaluated using Kappa statistics.

RESULTS

The intra-rater reliability for UIV and LIV was 0.657 and 0.612 between rounds 1 and 2, 0.634 and 0.692 between rounds 1 and 3, and 0.659 and 0.638 between rounds 2 and 3, respectively, which indicated substantial agreement between rounds. The inter-rater kappa reliabilities for UIV and LIV selection were 0.103 and 0.412 for round 1, 0.121 and 0.380 for round 2, and 0.125 and 0.368 for round 3, indicating slight to moderate agreement between raters.

CONCLUSIONS

Whether raters used either SB or TX radiography, or both in addition to PA and lateral standing radiographs, did not influence the decision making for UIV or LIV of AIS Lenke type 1 surgery.

Choosing the Distal Fusion Levels in Lenke Type 1 Adolescent Idiopathic Scoliosis: How Do the Existing Classifications and Recommendations Guide Us?

STUDY DESIGN

Retrospective cohort.

OBJECTIVE

(a) To compare the recommendations of Lenke and Peking Union Medical College (PUMC) classifications in choosing distal fusion levels in Lenke 1 adolescent idiopathic scoliosis (AIS) curves and (b) to analyze whether the variability in distal fusion levels influences treatment outcomes.

METHODS

Hospital records of Lenke 1 AIS patients operated for single stage, posterior-only deformity correction were analyzed. Distal fusion levels recommended by Lenke and PUMC classifications were calculated and were compared with the actual distal fusion levels. The study population was divided based on whether the actual distal fusion levels were in agreement, shorter or longer than those recommended by Lenke classification. Subgroup analysis of Lenke 1C curves was done. The groups were compared with regard to the following outcome measures: Cobb angle correction, postoperative sagittal vertical axis, postoperative C7 offset, and Scoliosis Research Society-22r (SRS-22r) score at 24 months.

RESULTS

The distal fusion levels recommended by the 2 classifications were in agreement in 92 of 104 cases. In all the cases with disparity, Lenke classification recommended shorter fusions than the PUMC classification. No statistically significant difference was observed in the outcome measures-whether the actual distal fusion levels were in agreement, shorter, or longer than those recommended by the Lenke classification or whether or not the recommendations for selective fusion of any of these classifications were adhered to.

CONCLUSION

Lenke classification can save fusion levels without compromising on treatment outcomes when compared with PUMC classification. Variability in choice of distal fusion levels is not clinically significant at 24-month follow-up.

Impact of pelvic obliquity on coronal alignment in patients with adolescent idiopathic scoliosis

STUDY DESIGN

Retrospective study.

OBJECTIVE

To investigate pelvic obliquity prevalence and its coronal alignment effects in patients with adolescent idiopathic scoliosis (AIS). Pelvic obliquity observed on standing radiographs could affect coronal alignment; however, its incidence or relationship with coronal alignment remains unclear.

METHODS

Data of 141 patients with AIS (11 men, 130 women; mean age, 14.2 years) were retrospectively analyzed. Pelvic obliquity (pelvic coronal obliquity angle (PCOA) value ≥ 3°) was evaluated on standing radiographs; PCOAs were classified into two groups: right-upward and left-upward. Cobb angle and flexibility of proximal thoracic, main thoracic, thoracolumbar/lumbar curve, radiographic shoulder height (RSH), L4 tilt, and coronal balance were measured. Iliac crest and femoral head height differences were measured to evaluate correlations between pelvic obliquity and leg length discrepancy. We compared patients with and without pelvic obliquity according to Lenke classifications.

RESULTS

Among 141 patients, 33 (23%) showed pelvic obliquity: 12 were type 1, 3 were type 2, 1 was type 3, 13 were type 5, and 4 were type 6. Right-upward obliquity was observed in 25 patients (76%), all B or C curves; left-upward obliquity was observed in 8 (24%) and most were lumbar modifier A curves. PCOA and iliac crest height difference correlated with femoral head height difference. Among patients with Lenke type 1 with and without pelvic obliquity, those with right-upward pelvic obliquity showed significantly greater absolute RSH values, while those with left-upward pelvic obliquity showed significant smaller absolute RSH values. Among patients with Lenke types 5 and 6, those with pelvic obliquity showed significantly greater L4 tilt absolute values.

CONCLUSIONS

Pelvic obliquity was frequently observed in patients with AIS, especially lumbar scoliosis. Right-upward pelvic obliquity influenced by lumbar curves promoted shoulder imbalance and left-upward obliquity compensated for shoulder imbalance in Lenke type 1.

LEVEL OF EVIDENCE

Level 3.

Association Between Vertebral Rotation Pattern and Curve Morphology in Adolescent Idiopathic Scoliosis

OBJECTIVES

Although rotation is an important aspect of pathogenesis of adolescent idiopathic scoliosis (AIS), there are no studies demonstrating rotation pattern by directly measuring rotation angle in computed tomography (CT) images. The present retrospective comparative radiographic study was conducted to evaluate the rotation pattern in AIS and its relation to curve morphology and Lenke classification.

METHODS

The study included 245 patients diagnosed with AIS and evaluated with a preoperative CT scan. Rotation angle of each vertebrae was measured using reconstructed axial CT images. Lenke classification of scoliosis curvature, Cobb angle, location of apical vertebra and end vertebra, level of most rotated vertebra (MRV), and rotational shift vertebra were recorded. Student's t-test, analysis of variance test, and correlation analysis were performed to identify the characteristics of rotation patterns in each Lenke type.

RESULTS

The rotation angle of MRV was significantly correlated with the Cobb angle. The level of MRV and rotation angle of MRV demonstrated significant correlation in both the main thoracic curve and lumbar curve, which signified that MRV located near the thoracolumbar junction is related to more severe rotational deformity. The level of MRV was also significantly correlated to the Cobb angle in both the main thoracic curve and lumbar curve. The types of structural curves proximal to major structural curves, such as type 2, 4, and 6, demonstrated more severe rotational and coronal plane deformities as compared with types 1, 3, and 5.

CONCLUSIONS

Curves with different Lenke types demonstrated different levels of MRV and severity of rotation. The results suggest that different center levels of rotation, signified by MRV, are a factor determining curve morphology. The findings would be a basis of connection between 2-dimensional classification and transverse plane deformity.

An ontology-based module of the information system ScolioMedIS for 3D digital diagnosis of adolescent scoliosis

BACKGROUND AND OBJECTIVE

Conventional information systems are built on top of a relational database. The main weakness of these systems is impossibility to define stable data schema ahead when the knowledge of the system is evolving and dynamic. The widely accepted alternatives to relational databases are ontologies that can be used for designing information systems. Many research papers describe various methods for improving reliability and precision in generating the type of the Lenke classification based on the image processing techniques or a computer program, but all of them require radiograph images. The main objective of this paper is to demonstrate the development of an ontology-based module of the information system ScolioMedIS for adolescent idiopathic scoliosis (AIS) diagnosis and monitoring, which uses optical 3D methods to determine the Lenke classification of AIS and to avoid harmful effects of traditional radiation diagnosis.

METHODS

For creating an ontology-based module of the ScolioMedIS we used the following steps: specification, conceptualization, formalization and implementation. In the specification and conceptualization phase we performed data collection and analysis to define domain, concepts and relationships for ontology design. In the formalization and implementation stage we developed the OBR-Scolio ontology and the ontology-based module of the ScolioMedIS. The module employs the Protégé-OWL API, as a collection of Java interfaces for the OBR-Scolio ontology, which enables the creating, deleting, and editing of the basic elements of the OBR-Scolio ontology, as well as the querying of the ontology.

RESULTS

The ontology-based module of ScolioMedIS is tested on the datasets of 20 female and 15 male patients with AIS between the ages of 11 and 18, to categorize spinal curvatures and to automatically generate statistical indicators about the frequency of the basic spinal curvatures, degree of progression or regression of deformity and statistical indicators about curvature characteristics according to the Lenke classification system and Lenke scoliosis types. Results are then compared with analysis of the Lenke classification of 315 observed patients, performed using traditional radiation techniques.

CONCLUSIONS

This part of the system allows continuous monitoring of the progression/regression of spinal curvatures for each registered patient, which may provide a better management of scoliosis (diagnosis and treatment).

Three-Dimensional Radiographic Analysis of Two Distinct Lenke 1A Curve Patterns

STUDY DESIGN

Retrospective review of three-dimensional (3D) imaging from a multicenter database of surgically treated adolescent idiopathic scoliosis (AIS) patients.

OBJECTIVE

To use 3D analysis software to compare Lenke 1AR and 1AL curves in the coronal, sagittal, and axial planes.

BACKGROUND DATA

The Lenke 1AR and AL curve patterns have been shown to be two distinct curve types, with 1AL curves being more likely to add on after fusion. Analysis in 3D may help define some of the intricacies of these two curves.

METHODS

Ninety-four AIS patients with Lenke 1A curves and upright biplanar scanned radiographs were reviewed. Analysis was performed using 3D reconstruction software to evaluate the 3D coronal, sagittal and axial plane deformities. Coronal L4 tilt was used to distinguish between the two curve patterns.

RESULTS

The main thoracic Cobb was not significantly different between the AR (n = 43) and AL (n = 51) curves (52° ± 8° vs. 50° ± 5°; p = .25). The thoracolumbar/lumbar Cobb was significantly smaller in the AR curves (28° ± 8° vs. 32° ± 7°; p = .02). In the sagittal plane, T5-T12 kyphosis and T12-S1 lordosis were not significantly different (p >.2); however, the T10-L2 alignment was significantly more lordotic in the AR curves (11° ± 8° vs. 4° ± 10° lordosis; p <.001). In the axial plane, thoracic apical rotation was significantly greater in AR curves (21° ± 6° vs. 14° ± 6°; p <.001) and lumbar apical rotation was significantly smaller in AR curves (1° ± 5° vs. 6° ± 5°; p <.001).

CONCLUSION

3D spinal analysis demonstrates that 1AR and AL curves are distinctly different in all three planes. Although the treatment-based Lenke classification system combines these two curve patterns into one curve type, the 3D assessment suggests there are clear features that differentiate these curve patterns. The differing features of the nonstructural lumbar curves may help define the variance in fusion level selection and risk of adding-on for these two curve patterns.

LEVEL OF EVIDENCE

Level II, prognostic.

Selecting the lowest instrumented vertebra in adolescent idiopathic scoliosis: Comparison of the Lenke, Suk, and Dubousset criteria

BACKGROUND

Selection of the lowest instrumented vertebra (LIV) in patients undergoing selective fusion for Lenke type 1 or 2 adolescent idiopathic scoliosis (AIS) varies widely across centres around the world.

HYPOTHESIS

Lenke, Suk, and Dubousset criteria show moderate agreement for LIV selection.

METHODS

Sixty-eight patients with Lenke type 1 or 2 AIS managed by selective posterior fusion and followed-up for at least 2 years were included in a retrospective observational study. Agreement among Lenke, Suk, and Dubousset criteria for LIV selection was assessed. For surgery, the LIV was selected based on Dubousset criteria. Retrospectively, in each patient, the LIV selected by the Lenke and Suk criteria sets was identified on the preoperative images. The patients were then divided into two groups based on whether the Dubousset LIV was identical versus more distal than the LIV identified retrospectively by the Lenke or Suk criteria. The primary evaluation criterion was coronal balance.

RESULTS

The LIVs selected by the Lenke, Suk, and Dubousset criteria were identical in 57% of cases. The LIV selected by the Dubousset criteria were identical to that selected by the Lenke or Suk criteria in 70% of patients. No significant between-group differences were found for any of the evaluation criteria assessed preoperatively, postoperatively, or at last follow-up.

DISCUSSION

Agreement among the Lenke, Suk, and Dubousset criteria was moderate, confirming the working hypothesis. No coronal malalignment developed in the patients whose actual LIV was distal to the LIV selected by the Lenke or Suk criteria, supporting the validity of Dubousset criteria for LIV selection. When selecting the LIV, all three criteria sets should be assessed. The LIV is the vertebra selected by all three if they agree or by the Dubousset criteria if they do not.

LEVEL OF EVIDENCE

IV, retrospective cohort study.

Three-dimensional reconstructions of Lenke 1A curves

Background

Scoliosis is a 3D deformity that can be reconstructed through 2D antero-posterior and lateral radiographs, which provide an upper view of the deformed spine as well as regional planes matching all vertebrae of elective plane for each curve. The objective of this study is to explore whether all idiopathic scoliosis classified Lenke 1A have the same 3D representation made with regional planes.

Methods

All patients treated for idiopathic thoracic scoliosis during the growth period and classified Lenke 1A were included in this study conducted in the pediatric spinal orthopedic department of Centre des Massues. A photogrammetric technique was used to obtain a 3D reconstruction, from regional planes identified on radiographs made with the EOS system. Three regional planes are usually identified in asymptomatic spines: lumbar, dorsal, and cervical—none of them presenting rotation. In the studied group, the number of planes, the rotation, and the limit vertebrae of each plane were looked for.

Results

Sixty-three patients were included (47 girls and 16 boys, mean age 11.3 years). The Cobb angle was meanly 36.5°. The scoliosis was reconstructed with three regional planes (57%) or four ones (43%, with the thoracic plane divided into two planes). Maximal rotation was found in the thoracic plane, especially when scoliosis was represented with four regional planes. The transition between planes 2 and 3 was mainly located between the fourth and sixth dorsal vertebrae.

Conclusion

The use of an arbitrary regional plane representation of a 3D shape leads to conclude that there are two types of Lenke 1A scoliosis, which should be taken into account for designing the brace.

The Spinopelvic Geometry in Different Lenke Curve Types of Adolescent Idiopathic Scoliosis

BACKGROUND

The Lenke classification is well established in differentiation of curve types in adolescent idiopathic scoliosis (AIS) and guides selection of fusion levels. However, to date, it has neglected the spinopelvic parameters that have been associated with compensatory mechanisms in balancing the human erect posture and adjacent segment problems after spinal fusion. The aim of this study was to investigate spinopelvic parameters in different types of AIS curves.

MATERIAL AND METHODS

Preoperative whole-spine radiographs from 100 patients with AIS were reviewed and the curves were classified according to Lenke. In addition, sagittal spinopelvic parameters (pelvic incidence, sacral slope, pelvic tilt) were measured and compared between different curve types and to normal population values.

RESULTS

The spinopelvic balance was not statistically distinguishable in different Lenke curve types. Slight differences of the spinopelvic balance, compared with normal population values, were found in AIS Lenke Type 5 and 6 curves (major curve at the lumbar/thoracolumbar region) with a pelvic incidence of 44° ± 8° (norm 49°), sacral slope of 34° ± 7° (norm 41°), and pelvic tilt of 10° ± 7° (norm 8°).

CONCLUSION

Overall, the variances of spinopelvic parameters in different AIS curve types do not seem statistically large enough for a potential clinical relevance. However, the sacrum is more verticalized in AIS curves with major curves located in the lumbar/thoracolumbar region. It remains to be investigated whether such a verticalized sacrum might be a compensatory mechanism to keep the whole spine balanced and if it reverses with correction of the scoliosis.

Thoracic and lumbar vertebrae morphology in Lenke type 1 female adolescent idiopathic scoliosis patients

BACKGROUND

Pedicle screws are widely used in adolescent idiopathic scoliosis (AIS) surgeries. Pedicle screw malposition may lead to serious vascular and neurologic complications. Knowledge of the morphometric anatomy of the thoracic and lumbar vertebrae is essential for the surgeon while implanting pedicle screws. It has been reported that there is a reduction of pedicle width at the concavity of the curve in AIS patients. However, it is unclear if gender plays a role in this pedicle width pattern. The goal of this study is to assess the vertebrae morphology in a more homogeneous group of AIS patients - female patients with Lenke type 1 curve.

METHODS

The thoracic and lumbar vertebra and pedicle morphometry of 17 consecutive Lenke type 1 female AIS patients was analyzed based on 1mm fine cut CT scans. Morphometric anatomy of 539 pedicles from T1 to L5 was studied. Measurements included pedicle length, chord length, transverse pedicle width, transverse pedicle angle and vertebral rotation angle.

RESULTS

The mean age of the patients was 14 years old (range 12-18). The mean Cobb angle was 56° (range 43° -88°) and the mean angle of vertebral rotation varied between 4-13.8°. The apical vertebra was between T7 and T11. The transverse pedicle width was significantly smaller (p < 0.05) on the concave side in the apical region of the thoracic spine (T7 and T8), measuring between 2.1-2.2 mm on the concave side and 2.7-3.1 mm on the convex side. Meanwhile, in some upper thoracic vertebrae (T3, T4, T5), the width was significantly bigger (p < 0.05) on the concave side than on the convex side, measuring between 2.8-4 mm on the concave side and 1.8-2.4 mm on the convex side. In the lumbar spine, the width varied between 4.1-9.9 mm without significant differences between the concave and convex sides (p > 0.05). The pedicle length varied between 15.4-28.7 mm and was significantly smaller (p < 0.05) on the concave side at T4, T5 and L2. The chord length was shortest at T1, measuring 32.4 mm and increased gradually to 54.3 mm at L3 but no statistical difference was found between the concave and convex sides. The transverse pedicle angle varied between 11.8° and 35° and was significantly bigger on the concave side at T7 and on the convex side at L1 (p < 0.05).

CONCLUSIONS

The vertebrae morphology in Lenke type 1 female AIS patients is substantially different from the vertebrae in normal spines especially at the apex and in the upper thoracic region. This is consistent with some previous reports which did not distinguish between male and female patients. Our findings suggest that gender does not play a major role in the vertebrae morphology pattern of AIS patients. Furthermore, recognizing this pattern is critical in order to optimize pedicle screw instrumentation and may allow for some leeway adjustments in the pedicle screw trajectory regardless of the methods of implantation.

Comparison of Typical Thoracic Curves and Atypical Thoracic Curves Within the Lenke 1 Classification

STUDY DESIGN

Retrospective study.

OBJECTIVES

To examine the characteristics of Lenke type 1 curves based on the level of the apical vertebra.

SUMMARY OF BACKGROUND DATA

The Lenke classification is the most used system for adolescent idiopathic scoliosis, with approximately 50% of the curves falling into the Lenke 1 curve type category.

METHODS

A total of 611 Lenke 1 curves in a prospectively collected multicenter adolescent idiopathic scoliosis study were analyzed. Minimum follow-up was ≥ 2 years. Curves were subdivided into 3 groups according to their apex: the typical Lenke 1 curve group included apices from T7/8 to T10 (511 patients), the proximal group included apices from T4 to T7 (45 patients), and the distal group included apices from T10/11 to T11/12 (50 patients). Preoperative and postoperative radiographic and clinical outcomes were compared among the 3 groups.

RESULTS

The proximal and distal groups included significantly more left thoracic curves (proximal: 29%; typical: 1.8%; distal: 19%; p < .01). Flexibility of the main thoracic curve was significantly different among the 3 groups (proximal: 32% ± 17%; typical: 46% ± 18%; distal: 57% ± 18%; p < .001). The distal group included significantly more "A" lumbar modifiers (proximal: 29%; typical: 53%; distal: 96%; p < .01) and had curves characteristics similar to King type 4 curves (L4 tilted to the right: Lenke 1AR). The average lowest instrumented vertebra was significantly lower in the distal group (proximal: T12; typical: L1; distal: L2; p < .01). The proximal group had significantly greater thoracic kyphosis (proximal: 30° ± 18°; typical: 20° ± 13°, distal: 20° ± 10°; p < .001) and more fusion segments (proximal: 10; typical: 9; distal: 9; p < .03).

CONCLUSIONS

Curves categorized as Lenke 1 curves were less homogeneous than expected. Using only the Lenke type 1 designation to define a study population may introduce unintended bias to the study design.

Artificial neural networks assessing adolescent idiopathic scoliosis: comparison with Lenke classification

BACKGROUND CONTEXT

Variability in classifying and selecting levels of fusion in adolescent idiopathic scoliosis (AIS) has been repeatedly documented. Several computer algorithms have been used to classify AIS based on the geometrical features, but none have attempted to analyze its treatment patterns.

PURPOSE

To use self-organizing maps (SOM), a kind of artificial neural networks, to reliably classify AIS cases from a large database. To analyze surgeon's treatment pattern in selecting curve regions to fuse in AIS using Lenke classification and SOM.

STUDY DESIGN

This is a technical concept article on the possibility and benefits of using neural networks to classify AIS and a retrospective analysis of AIS curve regions selected for fusion.

PATIENT SAMPLE

A total of 1,776 patients surgically treated for AIS were prospectively enrolled in a multicentric database. Cobb angles were measured on AIS patient spine radiographies, and patients were classified according to Lenke classification.

OUTCOME MEASURES

For each patient in the database, surgical approach and levels of fusion selected by the treating surgeon were recorded.

METHODS

A Kohonen SOM was generated using 1,776 surgically treated AIS cases. The quality of the SOM was tested using topological error. Percentages of prediction of fusion based on Lenke classification for each patient in the database and for each node in the SOM were calculated. Lenke curve types, treatment pattern, and kappa statistics for agreement between fusion realized and fusion recommended by Lenke classification were plotted on each node of the map.

RESULTS

The topographic error for the SOM generated was 0.02, which demonstrates high accuracy. The SOM differentiates clear clusters of curve type nodes on the map. The SOM also shows epicenters for main thoracic, double thoracic, and thoracolumbar/lumbar curve types and transition zones between clusters. When cases are taken individually, Lenke classification predicted curve regions fused by the surgeon in 46% of cases. When those cases are reorganized by the SOM into nodes, Lenke classification predicted the curve regions to fuse in 82% of the nodes. Agreement with Lenke classification principles was high in epicenters for curve types 1, 2, and 5, moderate in cluster for curve types 3, 4, and 6, and low in transition zones between curve types.

CONCLUSIONS

An AIS SOM with high accuracy was successfully generated. Lenke classification principles are followed in 46% of the cases but in 82% of the nodes on the SOM. The SOM highlights the tendency of surgeons to follow Lenke classification principles for similar curves on the SOM. Self-organizing map classification of AIS could be valuable to surgeons because it bypasses the limitations imposed by rigid classification such as cutoff values on Cobb angle to define curve types. It can extract similar cases from large databases to analyze and guide treatment.

Are Thoracic Curves With a Low Apex (T11 or T11/T12) Really Thoracic Curves?

STUDY DESIGN

Multicenter; review of prospectively collected data.

OBJECTIVES

To determine whether curves with an apex at T11 or T11/T12 represent a combination of thoracic and thoracolumbar curves or constitute their own class of curves.

SUMMARY OF BACKGROUND DATA

Curves with an apex at T11 and T11/T12 are classified by the Scoliosis Research Society and Lenke classification as thoracic curves.

METHODS

We reviewed 1,835 adolescent idiopathic scoliosis cases from a multicenter database. Based on the most common curve apex levels (thoracic [TH] curves, T9; thoracolumbar/lumbar [TL/L] curves, L1), we created 2 comparison groups that represented classic examples of TH and TL/L curves. We identified 66 cases with an apex at T11-T11/T12 and compared them with 320 cases with a T9 apex, and 126 cases with an L1 apex. We compared characteristics of these 3 groups (p < .05) and completed a radiographic review of the T11-T11/T12 group.

RESULTS

The curve direction of the 3 groups was significantly different (TL/L, 84% left; TH, 3% left; and T11-T11/T12, 16% left) (p < .001). The mean number of vertebrae in curves for the T11-T11/T12 group (7.1 ± 1.2) fell between the value for the TL/L (5.7 ± 0.8) and TH (7.3 ± 1.0) groups. The T11-T11/T12 curves had a greater trunk shift than TL/L (p = .002) and TH (p = .011) curves. There was no difference among the 3 groups in terms of major curve Cobb magnitude (p = .09) or age at surgery (p = .76). Radiographic review of the T11-T11/T12 curves revealed 3 curve patterns: 21 long single curves (32%), 28 short single curves (42%), and 17 double thoracic curves (26%).

CONCLUSIONS

We suggest caution in lumping curves with an apex at T11 or T11-12 disc together with other thoracic apices in studies involving primary thoracic curves, because some of these curves have features much more typical of thoracolumbar curves.