Centroid method: reliable method to determine the coronal curvature of scoliosis: a case control study comparing with the Cobb method

Estimation of plane of maximum curvature for the patients with adolescent idiopathic scoliosis via a purpose-design computational method

PURPOSE

The coronal Cobb angle is commonly used for assessing the adolescent idiopathic scoliosis (AIS); however, it may underestimate the severity of AIS while the plane of maximum curvature (PMC) could be a promising descriptor for three-dimensional assessment of AIS. This study aimed to develop a computational method (CM) for estimating the PMC based on the coronal and sagittal images of the spine, and to verify the results with computed tomography (CT).

METHODS

Twenty-eight thoracic and 24 lumbar curves from 30 subjects with AIS were analysed. For the CM, PMC was estimated via identifying the two corner points at the superior endplate of upper-end vertebra and the inferior endplate of lower-end vertebra in the coronal and sagittal CT images separately (eight corner points in total). For the CT, PMC was determined through rotating the spine images axially until the maximum Cobb angle was found. Intraclass correlation coefficient (ICC), Bland-Altman method and linear regression analysis were used for the statistical analyses.

RESULTS

The high ICC values (intra- > 0.91; inter- > 0.84) suggested very good intra- and inter-rater reliability of the CM in PMC estimation. The high ICC values (> 0.91) and assessment of Bland-Altman method demonstrated a good agreement between the PMC acquired using the CM and CT. The generated linear regression equations (R² > 0.69) could allow to estimate the PMC (originally measured through the CT) via the CM.

CONCLUSION

The developed computational method could estimate reliable and valid PMC for the patients with AIS, and become feasible for three-dimensional assessment of AIS.

LEVEL OF EVIDENCE

III.

A Practical Study of Diagnostic Accuracy: Scoliosis Screenings of Middle School Students by a Trained Nurse With a Smartphone Versus a Spine Surgeon With a Scoliometer

STUDY DESIGN

Cross-sectional.

OBJECTIVE

This study aimed to assess the accuracy of smartphone-aided diagnosis of scoliosis by a trained nurse compared with scoliometer-based diagnosis by a spine surgeon.

SUMMARY OF BACKGROUND DATA

Many assessments have been developed to estimate the reliability of smartphone-aided measurements in diagnosing scoliosis. However, clinical studies assessing the accuracy of smartphone-aided diagnosis with radiographs or scoliometers are scarce.

METHODS

A total of 2702 grade 7 students (mean age 13.56 yrs, range 13-15) at 10 middle schools were first screened with a smartphone by a trained nurse from the orthopedics department. Approximately half a year later, most of the students underwent a chest x-ray examination as part of a compulsory medical examination. Students with suspicious findings in either the first screen or the chest x-ray were recommended to a scoliosis clinic for single-blind tests, such as a forward bending test (FBT) and an analysis of the angle of trunk rotation (ATR) with a scoliometer, performed by an experienced spine surgeon. Finally, the Cobb method was conducted with full-spine radiographs to serve as the gold standard.

RESULTS

The agreement between the first screening by the nurse and the second test by the spine surgeon was low in cases with a Cobb angle <10° (κ = 0.128 [0.04-0.22], P = 0.035) and fair in cases with a Cobb angle >10° (κ = 0.349 [0.19-0.50], P < 0.001). The results of receiver operating characteristic (ROC) curve analysis also suggested that these two tests were similar in their ability to diagnose scoliosis. However, when the Cobb angle cutoff was adjusted to 15°, the latter had markedly better diagnostic ability than the former. Overall, the sensitivity of the smartphone screening was not acceptable for recognizing scoliosis.

CONCLUSION

This study revealed that smartphone-aided screening for scoliosis is risky.

LEVEL OF EVIDENCE

3.

A computerized method for evaluating scoliotic deformities using elliptical pattern recognition in X-ray spine images

BACKGROUND AND OBJECTIVE

Several studies have evaluated the reproducibility of the Cobb angle for measuring the degree of scoliotic deformities from X-ray spine images, and proposed different geometric models for describing the spinal curvature. The ellipse was shown to be an adequate geometric form, but was not yet applied for the identification and quantification of scoliotic curvatures. The purpose of this study is therefore to propose and validate a novel computerized methodology for the detection of elliptical patterns from X-ray images to evaluate the extent of the underlying scoliotic deformity.

METHODS

For anteroposterior each X-ray spine image, the spine curve is first reconstructed from vertebral centroids. The ellipse that best fits to the obtained spine curve is the found within a least square and genetic algorithm optimization framework. The geometric parameters of the resulting best fit ellipse are finally used to define an index that quantifies the spinal curvature.

RESULTS

The proposed methodology was validated on three synthetic images and then successfully applied to 20 clinical anteroposterior X-ray spine images of patients with a different degree of scoliotic deformity, with the resulting maximal relative error of 3% for the synthetic images and an overall error of 0.5 ± 0.4 mm (mean ± standard deviation) for the clinical cases.

CONCLUSIONS

The results indicate that the proposed computerized methodology is able to reliably reproduce scoliotic curvatures using the geometric parameters of the underlying ellipses. In comparison to conventional approaches, the proposed methodology potentially produces less errors, requires a relatively low observer interaction, takes into account all vertebrae within the observed scoliotic deformity, and allows for both qualitative and quantitative evaluations that may complement the diagnosis, study and treatment of scoliosis.

Do Current Recommendations for Upper Instrumented Vertebra Predict Shoulder Imbalance? An Attempted Validation of Level Selection for Adolescent Idiopathic Scoliosis

BACKGROUND

Shoulder balance for adolescent idiopathic scoliosis (AIS) patients is associated with patient satisfaction and self-image. However, few validated systems exist for selecting the upper instrumented vertebra (UIV) post-surgical shoulder balance.

QUESTIONS/PURPOSES

The purpose is to examine the existing UIV selection criteria and correlate with post-surgical shoulder balance in AIS patients.

METHODS

Patients who underwent spinal fusion at age 10-18 years for AIS over a 6-year period were reviewed. All patients with a minimum of 1-year radiographic follow-up were included. Imbalance was determined to be radiographic shoulder height |RSH| ≥ 15 mm at latest follow-up. Three UIV selection methods were considered: Lenke, Ilharreborde, and Trobisch. A recommended UIV was determined using each method from pre-surgical radiographs. The recommended UIV for each method was compared to the actual UIV instrumented for all three methods; concordance between these levels was defined as "Correct" UIV selection, and discordance was defined as "Incorrect" selection.

RESULTS

One hundred seventy-one patients were included with 2.3 ± 1.1 year follow-up. For all methods, "Correct" UIV selection resulted in more shoulder imbalance than "Incorrect" UIV selection. Overall shoulder imbalance incidence was improved from 31.0% (53/171) to 15.2% (26/171). New shoulder imbalance incidence for patients with previously level shoulders was 8.8%.

CONCLUSIONS

We could not identify a set of UIV selection criteria that accurately predicted post-surgical shoulder balance. Further validated measures are needed in this area. The complexity of proximal thoracic curve correction is underscored in a case example, where shoulder imbalance occurred despite "Correct" UIV selection by all methods.

Reliability and Validity Study of Clinical Ultrasound Imaging on Lateral Curvature of Adolescent Idiopathic Scoliosis

Background

Non-ionizing radiation imaging assessment has been advocated for the patients with adolescent idiopathic scoliosis (AIS). As one of the radiation-free methods, ultrasound imaging has gained growing attention in scoliosis assessment over the past decade. The center of laminae (COL) method has been proposed to measure the spinal curvature in the coronal plane of ultrasound image. However, the reliability and validity of this ultrasound method have not been validated in the clinical setting.

Objectives

To evaluate the reliability and validity of clinical ultrasound imaging on lateral curvature measurements of AIS with their corresponding magnetic resonance imaging (MRI) measurements.

Methods

Thirty curves (ranged 10.2°–68.2°) from sixteen patients with AIS were eligible for this study. The ultrasound scan was performed using a 3-D ultrasound unit within the same morning of MRI examination. Two researchers were involved in data collection of these two examinations. The COL method was used to measure the coronal curvature in ultrasound image, compared with the Cobb method in MRI. The intra- and inter-rater reliability of the COL method was evaluated by intra-class correlation coefficient (ICC). The validity of this method was analyzed by paired Student’s t-test, Bland–Altman statistics and Pearson correlation coefficient. The level of significance was set as 0.05.

Results

The COL method showed high intra- and inter-rater reliabilities (both with ICC (2, K) >0.9, p<0.05) to measure the coronal curvature. Compared with Cobb method, COL method showed no significant difference (p<0.05) when measuring coronal curvature. Furthermore, Bland-Altman method demonstrated an agreement between these two methods, and Pearson’s correlation coefficient (r) was high (r>0.9, p<0.05).

Conclusion

The ultrasound imaging could provide a reliable and valid measurement of spinal curvature in the coronal plane using the COL method. Further research is needed to validate the proposed ultrasound measurement in larger clinical trial and to optimize the ultrasound scanning and measuring procedure.

Reliability of Coronal Curvature Measures in Premature Scoliosis: Comparison of 4 Methods Using Inverted Digital Luminescence Radiography

STUDY DESIGN

Observational study with 4 examiners.

OBJECTIVE

To improve the reproducibility and reliability of coronal curvature measurements.

SUMMARY OF BACKGROUND DATA

Several different methods have been developed to measure coronal curvature. Inverted digital luminescence radiography may be an additional technique to improve the reliability of measurement methods.

METHODS

A total of 114 whole spine posterior-anterior radiographs were collected to compare the reliability of 4 measurement methods used in early onset scoliosis. Each radiograph was measured twice by each of 4 examiners using 4 measurement methods with normal and inversion views. Data were analyzed to determine inter- and intraobserver reliabilities.

RESULTS

For all 4 methods, the overall intraobserver ICCs were greater than 0.926 and the interobserver ICCs were greater than 0.902, indicating excellent reliability. When intraobserver reliability was assessed in relation to scoliosis severity (<20°, 20°-40°, and >40°), the ICCs were higher for greater degrees of curvature; nonetheless, these values were in the excellent range for all methods and were higher for inversion (>0.908) than normal (>0.838) radiographs. For interobserver comparisons, the ICCs were better for the pedicle method (normal: >0.795, inversion: >0.831) than for the other methods (normal: >0.532, inversion: >0.590), with inverted digital luminescence radiography showing the higher ICCs.

CONCLUSION

Inversion radiographs showed high reliability and may increase reliability in cases of less deformed spines. In addition, the pedicle method led to higher ICCs for all measurements regardless of severity, especially on inversion radiographs.

LEVEL OF EVIDENCE

2.

Reliability analysis of shoulder balance measures: comparison of the 4 available methods

STUDY DESIGN

Observational study with 3 examiners.

OBJECTIVE

To compare the reliability of shoulder balance measurement methods.

SUMMARY OF BACKGROUND DATA

There are several measurement methods for shoulder balance. No reliability analysis has been performed despite the clinical importance of this measurement.

METHODS

Whole spine posteroanterior radiographs (n = 270) were collected to compare the reliability of the 4 shoulder balance measures in patients with adolescent idiopathic scoliosis. Each radiograph was measured twice by each of the 3 examiners using 4 measurement methods. The data were analyzed statistically to determine the inter- and intraobserver reliability.

RESULTS

Overall, the 4 radiographical methods showed an excellent intraclass correlation coefficient regardless of severity in intraobserver comparisons (>0.904). In addition, the mean absolute difference values in all methods were low and were comparatively similar (<1.73°). However, in interobserver comparisons, reliabilities were significantly decreased in the less severe radiographs, firstly on radiographical shoulder height measures (intraclass correlation coefficients >0.445, mean absolute difference <3.91°). However, the intraclass correlation coefficients in the coracoid height difference and clavicular angle methods were in the excellent range (>0.810 and >0.787, respectively) regardless of severity. In addition, the mean absolute difference values in the clavicular angle method were lower (<0.62°) than others.

CONCLUSION

The higher reliability of the clavicular angle and coracoid height difference methods indicate the clinical usefulness of these methods. Physicians should selectively use the shoulder balance measurement method clinically.

LEVEL OF EVIDENCE

3.

Centroid method: an alternative method of determining coronal curvature in scoliosis. A comparative study versus Cobb method in the degenerative spine

BACKGROUND CONTEXT

Although the Cobb method is considered the gold standard, the centroid method may offer a reasonable alternative in adult scoliosis because it has better inter- and intraobserver reliabilities in adolescent scoliosis.

PURPOSE

To compare the reliabilities of the Cobb and the centroid methods for measuring coronal curvature in degenerative scoliosis in older patients.

STUDY DESIGN

Observational study involving three examiners.

PATIENT SAMPLE

Sixty whole spine posteroanterior (PA) radiographs were collected.

OUTCOME MEASURES

Data were analyzed to determine inter- and intraobserver reliabilities.

METHODS

Sixty whole spine PA radiographs of patients older than 60 years were collected to compare the reliabilities of the centroid and the Cobb methods for measuring coronal curvature in degenerative scoliosis. Three examiners using both methods independently measured radiographs twice. Data were analyzed to determine inter- and intraobserver reliabilities.

RESULTS

Intraobserver comparisons of all the 60 radiographs revealed that inter- and intraclass correlation coefficients (ICCs) of the Cobb and the centroid methods were both excellent (greater than 0.979 vs. greater than 0.918), and mean absolute differences (MADs) were similar (less than 1.58 vs. less than 2.02). In interobserver comparisons, ICCs of the Cobb method were higher than that of the centroid method (greater than 0.922 vs. greater than 0.799), and the MADs of the Cobb method were lower than that of the centroid method (less than 2.91 vs. less than 4.84). Comparisons of radiographs subdivided by severity showed that the ICCs of the Cobb and the centroid methods were both excellent (greater than 0.819 vs. greater than 0.801), and their MADs were similar (less than 2.29 vs. less than 2.53) for intraobserver comparisons. Interobserver comparisons showed that ICCs and MADs were dependent on the severity of coronal curvature, and the ICCs of the Cobb method (greater than 0.698) were greater than that of the centroid method (greater than 0.507). Furthermore, MAD values for the Cobb method were lower than that for the centroid method (less than 3.59 vs. less than 6.07). Moreover, these results are contradictory to the previous study, which showed the higher reliability of the centroid method in measures of adolescent scoliosis.

CONCLUSIONS

In the present study, the reliability of the centroid method was found to be more susceptible to the severity of disease in older patients, despite its demonstrated greater reliability in adolescent scoliosis. Our findings show that the selective use of these two methods in old and young patients can increase the reliabilities of measurements made.

A comparative analysis of 4 curvature measurement methods in early-onset scoliosis

STUDY DESIGN

Observational study with 3 examiners.

OBJECTIVE

The aim of this study was to enhance the reproducibility and reliability of coronal curvature measurements in early-onset scoliosis.

SUMMARY OF BACKGROUND DATA

Previous reports show high variability of the Cobb method, especially on the measurement of the immature spine.

METHODS

A total of 115 whole-spine posteroanterior radiographs were collected to compare the reliability of the Cobb, lateral tangent, pedicle, and centroid methods in early-onset scoliosis. Radiographs were measured twice by each of the 3 examiners using the 4 measurement methods. Statistical analysis was performed to determine the inter- and intraobserver reliability.

RESULTS

In this study, total inter- and intraobserver inter- and intraclass correlation coefficients (ICCs) in 115 radiographs were excellent in all methods (ICCs >0.961). However, mean absolute differences (MADs) in the lateral tangent method were less than 3.78°, which was higher than other methods (MADs <2.95°). In analysis of different severity groups (<15°, 15°-30°, and >30°), total inter- and intraobserver ICCs gradually increased with increasing the severity of the deformity, whereas MADs of each severity group were similar despite their increased measurement scale. Particularly, interobserver ICCs and MADs of lateral tangent method were more than 0.474 and less than 3.76° with poor reliability, which showed high variability in the less deformed spine group (<15°). However, intraobserver ICCs and MADs of pedicle method were more than 0.853 and less than 2.61°, interobserver ICCs and MADs were more than 0.729 and less than 2.86° with excellent reliability, which showed constantly high reliability regardless of coronal curvature severity.

CONCLUSION

In this study, the pedicle method showed constantly higher ICCs and lower MAD values in the early-onset scoliosis regardless of severity. However, the other 3 methods showed lower ICCs and higher MAD values, which showed lowest reliability in the lateral tangent method. For improved treatment of early-onset scoliosis, we recommend the pedicle method for measuring curvature regardless of severity.