Validation of a new computer-assisted tool to measure spino-pelvic parameters

Accepted Manuscript Title: Validation of a new computer-assisted tool to measure spino-pelvic parameters Author: Renaud Lafage, Emmanuelle Ferrero, Jensen K. Henry, Vincent Challier, Bassel Diebo, Barthelemy Liabaud, Virginie Lafage, Frank Schwab PII: DOI: Reference:

S1529-9430(15)01373-X http://dx.doi.org/doi: 10.1016/j.spinee.2015.08.067 SPINEE 56564

To appear in:

The Spine Journal

Received date: Revised date: Accepted date:

6-5-2015 19-8-2015 27-8-2015

Please cite this article as: Renaud Lafage, Emmanuelle Ferrero, Jensen K. Henry, Vincent Challier, Bassel Diebo, Barthelemy Liabaud, Virginie Lafage, Frank Schwab, Validation of a new computer-assisted tool to measure spino-pelvic parameters, The Spine Journal (2015), http://dx.doi.org/doi: 10.1016/j.spinee.2015.08.067. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Title: Validation of a new computer-assisted tool to measure spino-pelvic parameters

2 3

Authors: Renaud Lafage, MS1; Emmanuelle Ferrero, MD1; Jensen K. Henry, BA1; Vincent

4

Challier, MD1; Bassel Diebo, MD1; Barthelemy Liabaud, MD1; Virginie Lafage, PhD1; Frank

5

Schwab, MD1

6 7

Affiliations:

8

1. NYU Hospital for Joint Diseases; Department of Orthopaedic Surgery; New York; NY

9 10

Corresponding Author:

11

Virginie Lafage, PhD

12

306 East 15th Street, 1F

13

New York NY 10003

14

Phone: (646) 794-8643

15

Fax: (646) 602-6926

16

Cell: (646) 912-5820

17 18 19 20

ABSTRACT

21 22

Background Context: Evaluation of sagittal alignment is essential in the operative treatment of

23

spine pathology, particularly adult spinal deformity (ASD). However, software applications for

24

detailed spino-pelvic analysis are usually complex and not applicable to routine clinical use 1 Page 1 of 23

1

Purpose: To validate a new clinician-friendly software (Surgimap) in the setting of ASD.

2

Study Design/Setting: Accuracy and inter- and intra-rater reliability of a spine measurement

3

software. Five users (2 experienced –spine surgeon, 3 novice – spine researched fellow)

4

independently performed each part of the study in two rounds with one week between

5

measurements.

6

Patient Sample: 50 ASD patients drawn from a prospective database.

7

Outcome Measures: Spinal, pelvic, and cervical measurements parameters (including pelvic tilt

8

[PT], pelvic incidence [PI], lumbar-pelvic mismatch [PI-LL], lumbar lordosis [LL], thoracic

9

kyphosis [TK], T1 spino-pelvic inclination [T1SPI], sagittal vertical axis [SVA], cervical

10

lordosis [CL]).

11

Methods: For the accuracy evaluation, 30 ASD patient radiographs were pre-marked for

12

anatomic landmarks. Each radiograph was measured twice with the new software (Surgimap);

13

measurements were compared to those from a previously validated software. For the reliability

14

and reproducibility evaluation, users measured 50 unmarked ASD radiographs in two rounds.

15

Intra-class correlation (ICC) and International Standardization Organization (ISO)

16

reproducibility values were calculated. Measurement time was recorded.

17

Results: Surgimap demonstrated excellent accuracy as assessed by the mean absolute difference

18

from validated measurements: PT 0.12°, PI 0.35°, LL 0.58°, PI-LL 0.46°, TK 5.25°, T1SPI

19

0.53°, SVA 2.04mm. The inter- and intra-observer reliability analysis revealed good to excellent

20

agreement for all parameters. The mean difference between rounds was <0.4° for PT, PI, LL, PI-

21

LL, and T1SPI and <0.3mm for SVA. For PT, PI, LL, PI-LL, TK, T1SPI, and SVA, the intra-

22

observer ICC values were all >0.93 and the inter-observer ICC were all >0.87. Parameters based

23

on point landmarks rather than endplate orientation had a better reliability (ICC≥0.95 vs.

2 Page 2 of 23

1

ICC≥0.84). The average time needed to perform a full spino-pelvic analysis with Surgimap was

2

75 seconds (+25).

3

Conclusions: Using this new software tool, a simple method for full spine analysis can be

4

performed quickly, accurately, and reliably. The proposed list of parameters offers quantitative

5

values of the spine and pelvis, setting the stage for proper pre-operative planning. The new

6

software tool provides an important bridge between clinical and research needs.

7

Keywords: computerized measurement; spino-pelvic alignment; spinal deformity; reliability;

8

accuracy; reproducibility

9 10 11

3 Page 3 of 23

1

INTRODUCTION

2 3

The importance of sagittal alignment analysis in adult spinal deformity (ASD) has long been

4

emphasized 1, and numerous studies have reported the deleterious impact of spino-pelvic

5

malalignment on patient-reported outcomes

6

an essential tool in treatment planning for ASD and can be used to calculate the correction

7

needed 4–6; inadequate treatment can result in poor post-operative alignment and patient-reported

8

outcomes 7,8.

2,3

. Accurate sagittal alignment analysis has become

9 10

Despite its importance, analysis of the sagittal plane and individualized surgical planning can be

11

an arduous practice. In response, dedicated software has emerged to simplify sagittal plane

12

assessment. Multiple forms of digital measurement software have been shown to be faster, more

13

accurate, more precise, and less variable than manual techniques

14

measurements provide a step forward in sagittal spino-pelvic evaluation, but their application

15

remains largely

16

importance of such software, less than half regularly use software to analyze and plan sagittal

17

realignment procedures 11.

9,10

. These computer-based

limited to research: though nearly 90% of spine surgeons recognize the

18 19

Realistically, even with computerized software, the process of spino-pelvic analysis on

20

radiographs has until recently involved several complex steps and many current software

21

applications are considered time-consuming, technically demanding, and unpractical. Most

22

existing dedicated spine software programs require extensive measurements in order to generate

23

spino-pelvic parameter values12,13. In one such case, the user must identify the femoral heads and

24

the superior and inferior endplates of every single vertebra in the spine in order to generate 4 Page 4 of 23

1

measurements; in addition, such programs have complicated interfaces and are often limited to

2

the computers in which they are installed. As a result, these programs are used primarily in the

3

research realm, but are difficult to implement for routine clinical use.

4 5

Surgimap (Nemaris Inc., New York, NY) is a dedicated spine measurement and surgical

6

planning software that was developed to be applicable in both the research and the clinical

7

realms14. Within one application, users can upload images and use a variety of tools for

8

measuring coronal and sagittal radiographs. By identifying key parameters on the radiograph

9

with 11 cursor clicks, an entire collection of spino-pelvic parameters can be automatically and

10

instantaneously generated. Parameters can be modified, shown to patients, and saved by the

11

clinician for future analysis. Finally, Surgimap can be run from either the desktop or a portable

12

storage device (such as a universal serial bus [USB]), allowing the entire database of images and

13

measurements to be carried with the physician to the office, the operating room, or any other

14

location, without the need of additional software or specific operating systems. Authors have

15

advocated the use of this platform based on its effective and clinically useful osteotomy

16

simulations and surgical planning15. The reliability of this software has been established in a

17

recent study

18

alignment tool has been incorporated into the software.

16

, but since that publication, a more sophisticated, yet user-friendly sagittal

19 20

The current study validates the accuracy and reliability of this specialized sagittal alignment tool.

21

By simplifying the sagittal plane analysis, the software can provide a more efficacious and

22

applicable method to routine clinical evaluation of important spino-pelvic parameters.

23

5 Page 5 of 23

1

MATERIALS AND METHODS

2 3

Patient Population

4

For this single-center study, subjects were selected from a database of adult spinal deformity

5

(ASD) patients. Institutional review board (IRB) approval was obtained before study initiation.

6

Subjects were enrolled if they were older than 18 years and met radiographic criteria for ASD:

7

coronal Cobb angle >20°, sagittal vertical axis >5cm, pelvic tilt >25°, or thoracic kyphosis >60°.

8

Patients were excluded if their deformity was due to malignant, neuromuscular, infectious, or

9

traumatic etiology. Both primary and revision patients undergoing surgical or non-surgical

10

treatment were included.

11 12

Image Protocol

13

Each patient underwent spinal radiographic evaluation with anteroposterior and lateral views.

14

Radiographs were either low dose, full-spine or full-body (head to feet) biplanar

15

stereoradiographic images (EOS imaging, Paris, France) or full-spine 36-inch films. The protocol

16

for image acquisition called for a weight-bearing free-standing position with arms flexed at 45˚,

17

fingertips on clavicles, to avoid superimposition with the spine 17. Patients refrained from using

18

assistive devices to obtain the most representative standing position. The digitally obtained

19

images were recorded on a picture archiving and communication system (PACS) in digital

20

imaging and communications in medicine (DICOM) format and then uploaded for measurement

21

in the designated software programs.

22

6 Page 6 of 23

1

Sagittal Spino-Pelvic Parameters

2

Spino-pelvic parameters measured are shown in Figure 1. Pelvic parameters were Pelvic

3

Incidence [PI], Pelvic Tilt [PT], Sacral Slope [SS]; thoracolumbar measurements included

4

Lumbar Lordosis [LL], PI-LL mismatch [PI-LL], and Thoracic kyphosis [TK, T4-T12]. Global

5

alignment parameters consisted of Sagittal vertical axis [SVA], T1 spino pelvic inclination

6

[T1SPi], T9 spino pelvic inclination [T9SPi], and T1 pelvic angle [TPA] 18. Cervical parameters

7

included Cervical lordosis [CL], T1 slope [T1S], T1 slope -CL mismatch [TS-CL], and C2-C7

8

SVA [cSVA].

9 10

Measurement Process

11

Two software programs were used for measurements. The first, SpineView (ENSAM ParisTech,

12

Paris, France), is a validated software that requires identification of the femoral heads and

13

superior and inferior endplates of all vertebrae in the spine

14

sagittal alignment tool was employed (Figure 2). The user first outlines the femoral heads with

15

two adjustable circles; second, the user marks 4 segments corresponding to 4 key vertebral

16

endplates (superior S1, superior L1, superior T1, and inferior C2). Once the landmarks are

17

identified, constrained spines are adjusted by the user to correctly overlay the cervical, thoracic,

18

and lumbar curvatures. Spino-pelvic parameters are automatically generated by the software

19

based on the demarcated landmarks and splines: direct parameters correspond to angles drawn by

20

the observer (PI, PT, SS, LL, C2C7, PI-LL), while indirect parameters correspond to angles that

21

the software estimates based on the splines. In the event of non-calibrated radiographs,

22

calibration was performed using the length of the upper endplate of L3 as 35mm 20. The software

23

allows the use to adjust for brightness, contrast, white balance and gamma level; the user can

19

. For Surgimap, the specialized

7 Page 7 of 23

1

also modify any of the landmarks after the endpoints and splines are placed to generate updated

2

measurements.

3 4

Surgimap Evaluation and Statistical Analysis

5 6

Part 1: Accuracy

7

The first component of this study tested the accuracy of the Surgimap tool compared to the gold-

8

standard, SpineView (Surgiview, Paris, France)

9

the true reference values, as SpineView is currently one of the most widely used software

21

. SpineView measurements were accepted as

22

10

programs in research and has shown excellent accuracy and reliability

. Thirty ASD patient

11

radiographs were selected from the ASD patient database and measured using SpineView; they

12

were then marked with key landmarks to eliminate user bias in the accuracy evaluation (Figure

13

3). Five observers (2 experienced – spine surgeon, 3 novice – spine research fellow) measured

14

each radiograph in the new software twice using the provided landmarks, with one week between

15

measurements. Measurements from Surgimap and SpineView were used to calculate the mean

16

absolute difference, minimum difference, and maximum difference. Bland-Altman plots were

17

constructed to assess the difference between Surgimap and SpineView for each parameter.

18 19

Part 2: Reliability

20

The second component of this study used 50 unmarked EOS full-body radiographs from the

21

same ASD database. The 5 observers measured each radiograph twice with one week between

22

rounds (Figure 4). Patients were deidentified and the order was randomly changed between the

23

two sets of measurement. Mean values, standard deviations, and inter-observer difference were

8 Page 8 of 23

1

calculated for each parameter. Inter-rater (between raters) and intra-rater (within raters)

2

reliability was assessed using intra-class correlation coefficients (ICC). The ICC expresses the

3

proportion of global variability due to the subjects’ variability. The following thresholds

4

represent the quality of ICC: > 0.90, excellent; 0.71 to 0.90, good; 0.51 to 0.70, fair; 0.25 to 0.49,

5

fair, and <0.25, poor.23 Inter-observer reproducibility was evaluated using the International

6

Organization for Standardization (ISO5725-2)

7

standard (5725-2) assesses trueness (the closeness of the sample mean and the accepted value)

8

and precision (the closeness of agreement between test results) to describe the accuracy of a

9

measurement method. Finally, the time required to complete each x-ray measurement was

24

, with an effect size of 1°

24

. The international

10

recorded.

11

All data were statistically analyzed using SPSS (Inc. Version 20.0; Armonk, NY: IBM Corp),

12

Stata software 13.0 (Statacorp, College Station, Texas), MATLAB (Version R2010a; The

13

MathWorks, Inc., Natick, Massachussetts), and Excel (Microsoft Excel 2013; Redmond,

14

Washington). Statistical significance was set at P<0.05.

15 16

RESULTS

17 18

Part 1: Accuracy

19

Mean values of key parameters as measured by SpineView were as follows: PT: 26.5° ± 14.3, PI:

20

62.2° ± 12.3, PI–LL: 10.4° ± 28.0, and SVA: 39.8mm ± 70. Mean values for the same

21

parameters as measured by Surgimap were PT: 26.5° ± 14.3; PI: 62.0° ± 112.3 PI – LL: 10.4° ±

22

27.9; SVA: 38.4mm ± 69. Mean absolute differences and standard deviations are shown in Table

23

1. The minimum differences for all angular parameters was less than or equal to 0.01°. The

24

pelvic parameters (PI, SS, PT), which were directly measured by anatomic landmarks, tended to 9 Page 9 of 23

1

have the smallest differences between the new software and SpineView. The largest differences

2

were in TK and the cervical parameters; excluding these, all measurements were within 1°.

3

Bland-Altman plots demonstrating the differences between the specialized software and

4

SpineView are shown in Figure 5.

5 6

Part 2: Reliability

7

Means and standard deviations for both rounds of the reliability analysis are shown in Table 2.

8

For all angular parameters excluding TK and the cervical measurements, there was less than 0.5°

9

difference between measurements. The greatest difference between Rounds 1 and 2 was in TK,

10

in which the mean difference was 2.1°. Both linear parameters (cervical SVA and SVA) had

11

mean absolute differences of less than 0.5 millimeters.

12 13

Inter- and intra-observer reliability analyses revealed good (ICC>0.7) or excellent agreement

14

(ICC >0.9) for all parameters (Table 3). All intra-observer ICC values were >0.9 (excellent),

15

except T1 slope and the related parameter T1S-CL. Intra-observer agreement was highest for the

16

global alignment parameters (ICC > 0.970) and the three pelvic parameters (ICC >0.940). Inter-

17

observer reliability was also excellent for all parameters (ICC>0.799), with highest values for

18

global parameters and lowest for cervical measurements. Inter-observer agreement was modestly

19

lower for parameters based on endplate orientation (SS, LL, TK, CL; ICC > 0.840) than

20

parameters based on endpoints and angles (PT, T1SPi, T9SPi, TPA; ICC > 0.970).

21 22

Inter-observer reproducibility is also shown in Table 3. Angular global alignment parameters had

23

excellent reproducibility, as shown by values of less than 2° (TPA, T1SPI, T9SPI). Pelvic

10 Page 10 of 23

1

parameter reproducibility values (all <4°) were lower than those for thoracolumbar and cervical

2

parameters (4-6.5°). Reproducibility for both cSVA and SVA was <5mm. Finally, the average

3

time needed to perform a full spino-pelvic analysis in Surgimap was 75 seconds (+25).

4 5 6

DISCUSSION

7 8

Sagittal plane analysis is critical in the evaluation and treatment of spinal pathology, as well as

9

the assessment of post-operative alignment success. Radiographic evaluation is crucial, as

10

radiographic measurements usually correlate well with patient-reported outcomes 25. While most

11

providers appreciate the importance of radiographic evaluation, there is still a gap between the

12

awareness of its utility and the use of it in daily practice. One of the main limitations of spine

13

measurement software is the practicality of software and the time spent placing measurements on

14

images for each patient. While previous digital measurement techniques have been validated

15

19,21

16

identification of anatomic landmarks. Thus, the need for a clinically relevant and efficient spine

17

measurement software is evident.

, they are often cumbersome, incorporating difficult user interfaces and requiring extensive

18 19

Several studies have assessed the reliability and accuracy of computer-based radiographic

20

measurements, but using only healthy or asymptomatic subjects 26. Similarly, many authors have

21

assessed measurement reliability, yet only in the coronal plane (i.e. Cobb angle), as this was one

22

of the first parameters used to assess spinal deformity

23

digitized radiographs are not susceptible to the sources of error inherent to manual methods,

24

such as physical irregularities in protractor, wide-diameter markings, and inconsistencies the

23,27

. Early studies have noted that

11 Page 11 of 23

1

type of protractor or marker used 27. Further studies found that digitized software programs were

2

also superior to manual methods in sagittal plane measurements, noting their accuracy, reduced

3

inter- and intra-observer variability, and faster measurement time, even among experienced

4

spinal surgeons

5

used in the current study (identification of 12 anatomic landmarks used to generate spino-pelvic

6

parameters) had superior reproducibility and repeatability when compared to paper

7

measurements

8

Similarly, previous versions of Surgimap have been validated in studies, but have not

9

incorporated the new innovative sagittal alignment tool.16

21

12

. Recently, Maillot et al reported a software program similar to the platform

. However, this study did not include an evaluation of the software’s accuracy.

10 11

Despite the growing body of research supporting sagittal plane analysis, there is a need for

12

greater clinical applicability of tools, as many spine surgeons admit that while recognizing the

13

importance of dedicated software, they do not regularly use it in clinical practice

14

specialized software in this study provides a clinician-friendly interface with instructions and

15

graphical representations of each measurement method. Schwab et al utilized the software to

16

emphasize the importance of the pelvis in the sagittal plane

17

used this tool to analyze focal, regional and global measurements in the coronal and sagittal

18

plane

19

individual parameters such as coronal Cobb angle, TK, LL, PI, SS and PT

20

Alba et al confirmed the excellent inter- and intra-observer reliability of this software in specific

21

pelvic measurements (PI, SS, PT) 32.

8,29,30

28

11

. The

. Subsequently, numerous studies

. Reliability and reproducibility of this software platform has been demonstrated for 31

. More recently,

22

12 Page 12 of 23

1

This study introduces a new comprehensive sagittal alignment tool to provide a bridge between

2

clinical and research needs. The tool allows Surgimap to be immensely time-efficient; the

3

average time needed to generate a complete measurement of the spine from one radiograph was

4

little more than one minute (75 seconds). Comparison with the gold standard was not possible

5

due to the absence of this information. Estimates based on user experiences at our institution

6

revealed an average time of 3 to 15 minutes to generate the list of parameters, depending of the

7

degree of experience and the complexity of the analysis. Furthermore, the proposed list of

8

parameters is not exhaustive, but offers quantitative values of the spine (including the cervical

9

region) and pelvis. This combination of factors provides a novel and practical tool for physicians,

10

setting the stage for improved analysis and pre-operative planning.

11 12

The current study assessed the accuracy of the new sagittal alignment measurement tool by using

13

a validated software (SpineView) as the basis for reference values

14

measurement error between the old and new software tools: the mean error for indirect

15

parameters (LL, TK, T9SPi) was only modestly larger (0.53-5.25°) than that of the direct

16

parameters, and the mean error for direct parameters PT, PI, and PI-LL was consistently <0.5°.

19,21

. There was minimal

17 18

Intra- and inter-observer reliability analyses using ICC revealed excellent agreement for all

19

parameters. Previous authors have reported that endplate measurements are less reliable than

20

point measurements 33. Accordingly, in the current study, the T1SPi, T9SPi, and PT angles were

21

the most repeatable, with values ≥0.98. Other studies in the literature have reported intra-

22

observer reliability between 0.96-0.98 for computerized measurements of coronal Cobb angles27.

13 Page 13 of 23

1

The software also demonstrated excellent inter-observer reproducibility, as demonstrated by the

2

ICC values between 0.799-0.995 and small ISO reproducibility values. The lowest inter-class

3

ICC value occurred when observers measured regional cervical alignment (T1-CL).This result

4

can be explained by the curved shape of the cervical endplates: in order to interpolate the curved

5

shape of the inferior endplate of C2, the user must use an approximate straight line. An error of

6

measurement in C2 Slope impact directly the cervical lordosis (CL), and thus the T1 Slope minus

7

CL parameter. Point measurements and angles (T1SPi, T9SPi, TPA; pelvic parameters)

8

demonstrated minimal variability (less than 2° for global alignment parameters and less than 4°

9

for pelvic parameters). Variability was moderately higher for curves, but still less than 7°.

10

Comparatively, Vidal et al and Ilharreborde et al reported maximal inter-observer variability

11

values of 5°, though their studies focused on sagittal cervical alignment in adolescent idiopathic

12

scoliosis (AIS)

13

adult patients with ASD and assessed different radiologic parameters from the cervical, thoracic,

14

lumbar, sacral, and pelvic regions that correlate with clinical outcomes. Thus, both the accuracy

15

and reliability of this software were confirmed in multiple measurements in a symptomatic

16

population.

9,34

. This study analyzed two-dimensional parameters of EOS radiographs from

17 18

Some limitations of this study should be noted. The experience of the observers and the quality

19

of the radiographs were not taken into account. Radiograph quality and observer experience both

20

represent factors that should have the effect to diminish inter-observer reliability. However,

21

Segev et al noted no significant effect of expertise level on the digital measurement of various

22

pediatric orthopaedic parameters, remarking that a software with clear graphics and depictions of

23

tools allows users of all skill levels to accurately identify anatomical landmarks

35

. Yet by

14 Page 14 of 23

1

choosing radiographs of variable quality and users of different experiences, this study

2

emphasizes the broad clinical utility of the software in an applicable patient population.

3 4 5

CONCLUSION

6 7

Findings from this study demonstrate the accuracy, intra- and inter-observer reliability, and

8

efficacy of a spine measurement software tool. While measurements based on points were more

9

reliable than endplate angle parameters, all parameters showed good to excellent validity and

10

reliability. The software allows for measurement of radiographs with ease and speed, providing a

11

vast array of opportunities for assessment of spinal pathology for physicians and researchers

12

alike. Thus, the current tool will help promote the clinical application of sagittal plane spino-

13

pelvic analysis, and fills the gap between advanced research programs and routine clinical

14

practice for the care of spinal pathologies.

15 16

REFERENCES

17

18 19 20

1.

Schwab FJ, Ungar B, Blondel B, et al. Scoliosis Research Society-Schwab adult spinal deformity classification: a validation study. Spine (Phila Pa 1976). 2012;37(12):10771082. doi:10.1097/BRS.0b013e31823e15e2.

21 22 23

2.

Lafage V, Schwab FJ, Patel A, Hawkinson N, Farcy J-P. Pelvic tilt and truncal inclination: two key radiographic parameters in the setting of adults with spinal deformity. Spine (Phila Pa 1976). 2009;34(17):E599-E606. doi:10.1097/BRS.0b013e3181aad219.

24 25 26

3.

Glassman SD, Bridwell K, Dimar JR, Horton W, Berven S, Schwab FJ. The impact of positive sagittal balance in adult spinal deformity. Spine (Phila Pa 1976). 2005;30(18):2024-2029. doi:10.1097/01.brs.0000179086.30449.96.

15 Page 15 of 23

1 2 3

4.

Moal B, Lafage VC, Maier SP, et al. Discrepancies in Pre-operative Planning and Operative Execution in the Correction of Sagittal Spinal Deformities. In: IMAST 21st International Meeting on Advanced Spine Techniques. San Francisco; 2014:196.

4 5 6 7

5.

Ondra SL, Marzouk S, Koski T, Silva F, Salehi S. Mathematical calculation of pedicle subtraction osteotomy size to allow precision correction of fixed sagittal deformity. Spine (Phila Pa 1976). 2006;31(25):E973-E979. doi:10.1097/01.brs.0000247950.02886.e5 00007632-200612010-00024 [pii].

8 9 10

6.

Le Huec J-CC, Charosky S, Barrey C, Rigal J, Aunoble S. Sagittal imbalance cascade for simple degenerative spine and consequences: algorithm of decision for appropriate treatment. Eur Spine J. 2011;20 Suppl 5:699-703. doi:10.1007/s00586-011-1938-8.

11 12 13

7.

Schwab FJ, Patel A, Shaffrey CI, et al. Sagittal realignment failures following pedicle subtraction osteotomy surgery: are we doing enough?: Clinical article. J Neurosurg Spine. 2012;16(6):539-546. doi:10.3171/2012.2.SPINE11120.

14 15 16

8.

Blondel B, Schwab FJ, Bess S, et al. Posterior global malalignment after osteotomy for sagittal plane deformity: it happens and here is why. Spine (Phila Pa 1976). 2013;38(7):E394-E401. doi:10.1097/BRS.0b013e3182872415.

17 18 19 20

9.

Ilharreborde B, Steffen JS, Nectoux E, et al. Angle measurement reproducibility using EOS three-dimensional reconstructions in adolescent idiopathic scoliosis treated by posterior instrumentation. Spine (Phila Pa 1976). 2011;36(20):E1306-E1313. doi:10.1097/BRS.0b013e3182293548.

21 22 23 24

10.

Mok JM, Berven SH, Diab M, Hackbarth M, Hu SS, Deviren V. Comparison of observer variation in conventional and three digital radiographic methods used in the evaluation of patients with adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2008;33(6):681-686. doi:10.1097/BRS.0b013e318166aa8d.

25 26 27 28

11.

Lafage V, Schwab FJ, Scheer JK, et al. Adult Spinal Deformity Surgeons are Unable to Accurately Predict Post-operative Spinal Alignment: Initial Analysis of a Three-Phase Study. In: International Meeting on Advanced Spine Techniques (IMAST); July 16-19. Valencia, Spain; 2014.

29 30 31

12.

Maillot C, Ferrero E, Fort D, Heyberger C, Le Huec J-C. Reproducibility and repeatability of a new computerized software for sagittal spinopelvic and scoliosis curvature radiologic measurements: Keops(®). Eur Spine J. 2015. doi:10.1007/s00586-015-3817-1.

32 33 34

13.

Champain S, Benchikh K, Nogier a., Mazel C, Guise J De, Skalli W. Validation of new clinical quantitative analysis software applicable in spine orthopaedic studies. Eur Spine J. 2006;15(6):982-991. doi:10.1007/s00586-005-0927-1.

16 Page 16 of 23

1 2 3

14.

Akbar M, Terran J, Ames CP, Lafage V, Schwab F. Use of Surgimap Spine in Sagittal Plane Analysis, Osteotomy Planning, and Correction Calculation. Neurosurg Clin N Am. 2013;24(2):163-172. doi:10.1016/j.nec.2012.12.007.

4 5 6

15.

Akbar M, Terran J, Ames CP, Lafage V, Schwab FJ. Use of Surgimap Spine in Sagittal Plane Analysis, Osteotomy Planning, and Correction Calculation. Neurosurg Clin N Am. 2013;24(2):163-172. doi:10.1016/j.nec.2012.12.007.

7 8 9

16.

Vila-Casademunt A, Pellisé F, Acaroglu E, et al. The reliability of sagittal pelvic parameters: the effect of lumbosacral instrumentation and measurement experience. Spine (Phila Pa 1976). 2015;40(4):E253-E258. doi:10.1097/BRS.0000000000000720.

10 11 12

17.

Faro FD, Marks MC, Pawelek J, Newton PO. Evaluation of a functional position for lateral radiograph acquisition in adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2004;29(20):2284-2289. http://www.ncbi.nlm.nih.gov/pubmed/15480143.

13 14 15

18.

Lafage V, Protopsaltis TS, Schwab FJ, et al. The T1 Pelvic Angle (TPA), a Novel Radiographic Parameter of Sagittal Deformity, Correlates Strongly with Clinical Measures of Disability. Spine J. 2013;13(9):S61. doi:10.1016/j.spinee.2013.07.173.

16 17 18

19.

Rajnics P, Pomero V, Templier A, Lavaste F, Illes T. Computer-assisted assessment of spinal sagittal plane radiographs. J Spinal Disord. 2001;14(2):135-142. doi:10.1097/00002517-200104000-00008.

19 20 21 22 23

20.

Zhou SH, McCarthy ID, McGregor a H, Coombs RR, Hughes SP. Geometrical dimensions of the lower lumbar vertebrae--analysis of data from digitised CT images. Eur Spine J. 2000;9(3):242-248. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3611390&tool=pmcentrez&re ndertype=abstract.

24 25 26

21.

Rillardon L, Levassor N, Guigui P, et al. [Validation of a tool to measure pelvic and spinal parameters of sagittal balance]. Rev Chir Orthop Reparatrice Appar Mot. 2003;89(3):218227. http://www.ncbi.nlm.nih.gov/pubmed/12844045. Accessed September 26, 2014.

27 28 29

22.

Kim CH, Chung CK, Hong HS, Kim EH, Kim MJ, Park BJ. Validation of a simple computerized tool for measuring spinal and pelvic parameters. J Neurosurg Spine. 2012;16(2):154-162. doi:10.3171/2011.10.SPINE11367.

30 31 32 33

23.

Hardesty CK, Aronson J, Aronson EA, et al. Interobserver variability using a commercially available system of archived digital radiography with integrated computerassisted measurements for scoliosis Cobb angles. J Pediatr Orthop. 2013;33(2):163-169. doi:10.1097/BPO.0b013e3182770bd3.

34 35 36

24.

NF ISO 5725-2. Application of Statistics. Accuracy (Trueness and Precision) of Measurement Methods and Results. Part 2: Basic Method for the Determination of Repeatability and Repro- Ducibility of a Standard Measurement Method.; 2002.

17 Page 17 of 23

1 2 3 4

25.

Glassman SD, Berven S, Bridwell K, Horton W, Dimar JR. Correlation of radiographic parameters and clinical symptoms in adult scoliosis. Spine (Phila Pa 1976). 2005;30(6):682-688. http://www.ncbi.nlm.nih.gov/pubmed/15770185. Accessed August 12, 2013.

5 6 7 8

26.

Troyanovich SJ, Cailliet R, Janik TJ, Harrison DD, Harrison DE. Radiographic mensuration characteristics of the sagittal lumbar spine from a normal population with a method to synthesize prior studies of lordosis. J Spinal Disord. 1997;10(5):380-386. http://www.ncbi.nlm.nih.gov/pubmed/9355053. Accessed August 23, 2013.

9 10 11 12

27.

Shea KG, Stevens PM, Nelson M, Smith JT, Masters KS, Yandow S. A comparison of manual versus computer-assisted radiographic measurement. Intraobserver measurement variability for Cobb angles. Spine (Phila Pa 1976). 1998;23(5):551-555. http://www.ncbi.nlm.nih.gov/pubmed/9530786. Accessed July 31, 2014.

13 14 15

28.

Schwab F, Lafage V, Patel A, Farcy J-P. Sagittal plane considerations and the pelvis in the adult patient. Spine (Phila Pa 1976). 2009;34(17):1828-1833. doi:10.1097/BRS.0b013e3181a13c08.

16 17 18

29.

Kamerlink JR, Errico T, Xavier S, et al. Major intraoperative neurologic monitoring deficits in consecutive pediatric and adult spinal deformity patients at one institution. Spine (Phila Pa 1976). 2010;35(2):240-245. doi:10.1097/BRS.0b013e3181c7c8f6.

19 20 21

30.

Zhu Z, Xu L, Zhu F, et al. Sagittal alignment of spine and pelvis in asymptomatic adults: norms in Chinese populations. Spine (Phila Pa 1976). 2014;39(1):E1-E6. doi:10.1097/BRS.0000000000000022.

22 23 24

31.

Wu W, Liang J, Du Y, et al. Reliability and reproducibility analysis of the Cobb angle and assessing sagittal plane by computer-assisted and manual measurement tools. BMC Musculoskelet Disord. 2014;15:33. doi:10.1186/1471-2474-15-33.

25 26 27

32.

Vila-Casademunt A, Pellisé F, Acaroglu E, et al. The Reliability of Sagittal Pelvic Parameters: The Effect of Lumbosacral Instrumentation and Measurement Experience. Spine (Phila Pa 1976). 2014;40(4):E253-E258. doi:10.1097/BRS.0000000000000720.

28 29 30 31

33.

Harrison DE, Harrison DD, Cailliet R, Troyanovich SJ, Janik TJ, Holland B. Cobb method or Harrison posterior tangent method: which to choose for lateral cervical radiographic analysis. Spine (Phila Pa 1976). 2000;25(16):2072-2078. http://www.ncbi.nlm.nih.gov/pubmed/10954638.

32 33 34

34.

Vidal C, Ilharreborde B, Azoulay R, Sebag G, Mazda K. Reliability of cervical lordosis and global sagittal spinal balance measurements in adolescent idiopathic scoliosis. Eur Spine J. 2013;22(6):1362-1367. doi:10.1007/s00586-013-2752-2.

35 36

35.

Segev E, Hemo Y, Wientroub S, et al. Intra- and interobserver reliability analysis of digital radiographic measurements for pediatric orthopedic parameters using a novel

18 Page 18 of 23

1 2

PACS integrated computer software program. J Child Orthop. 2010;4(4):331-341. doi:10.1007/s11832-010-0259-5.

3 4

19 Page 19 of 23

1

FIGURE LEGENDS

2 3

Figure 1 (a-c). Pelvic parameters measured were PI, PT, and SS. Regional spinal parameters

4

included PI-LL mismatch, LL, and TK. Global alignment was assessed linearly by SVA and the

5

angular measurements of T1SPI, T9SPI, and TPA. Cervical parameters were comprised of T1

6

slope, C2-C7 cervical lordosis, and C2-C7 SVA.

7 8

Figure 2. Order of measurements as performed by Surgimap. Measurements are generated with a

9

total of 6 steps and can automatically generate more than 20 spino-pelvic parameters based on

10

the user-identified landmarks.

11 12

Figure 3. Sample measurement for the accuracy analysis. Radiographs were premarked with

13

anatomic landmarks, as shown by the white squares. Observers used the Surgimap software

14

sagittal alignment tool to create measurements, as shown.

15 16

Figure 4. For the inter- and intra-rater reliability portion of the analysis, observers made

17

measurements on unmarked radiographs.

18 19

Figure 5. Bland-Altman plots describing the difference between SpineView and Surgimap

20

measurements for parameters of interest (PI, PT, TK, LL, T1SPI, SVA).

21

Table 1. Differences between SpineView measurements and Surgimap.

22 Mean Absolute Mean Absolute

Minimum

Maximum

Absolute

Absolute

Parameter Difference

Diff. SD

20 Page 20 of 23

Difference

Difference

PT (°)

0.12

0.12

0.00

0.88

PI (°)

0.35

0.30

0.01

2.19

SS (°)

0.32

0.25

0.00

1.43

LL (°)

0.43

0.37

0.00

2.19

PI-LL (°)

0.35

0.36

0.00

3.87

TK (°)

5.25

4.70

0.00

14.69

T9SPI (°)

0.53

0.59

0.00

5.30

T1SPI (°)

0.14

0.10

0.00

0.41

TPA (°)

0.18

0.16

0.00

0.92

T1S (°)

0.55

0.53

0.00

2.99

CL (°)

2.28

1.66

0.01

7.81

TS-CL (°)

2.21

1.54

0.00

7.11

cSVA (mm)

4.55

7.67

0.00

42.46

SVA(mm)

2.04

2.12

0.03

23.52

1

21 Page 21 of 23

1 2

Table 2: Comparison of the mean values, standard deviations, and absolute difference

3

between rounds 1 and 2 of Surgimap software measurements. Key spino-pelvic parameters

4

were measured twice in 50 patients.

5

Round 1

Round 1

Round 2

Round 2

Mean

SD

Mean

SD

Parameter

Mean

Mean

Absolute

Absolute

Difference

Diff. SD

PT (°)

24.9

12.7

25.3

12.8

0.4

2.5

PI (°)

59.1

15.7

59.0

15.5

0.1

5.4

SS (°)

34.2

11.7

33.7

11.9

0.4

4.6

LL (°)

43.7

19.4

43.5

18.5

0.4

8.0

PI-LL (°)

15.0

21.4

15.5

21.2

0.4

6.4

TL (°)

9.4

28.6

9.6

27.7

0.1

6.9

TK (°)

36.1

21.2

38.2

21.0

2.1

8.4

T9SPI (°)

10.1

8.8

10.3

8.9

0.2

1.3

T1SPI (°)

1.2

8.6

1.1

8.7

0.0

0.8

TPA (°)

23.7

14.6

24.2

14.8

0.4

2.5

T1S (°)

32.0

14.6

33.0

15.1

1.1

6.8

CL (°)

8.0

15.9

8.7

16.4

0.8

7.4

C2-C7 SVA (mm)

23.6

11.8

23.6

12.0

0.4

9.5

SVA C7-S1 (mm)

52.2

67.5

51.7

66.0

0.3

5.1

6

22 Page 22 of 23

1

Table 3: Intra-observer and inter-observer reliability, as demonstrated by intra-class

2

correlation coefficients (ICC) on a scale of 0-1; and reproducibility (in degrees or mm) as

3

assessed by the International Standardization Organization.

4 5 Parameter

Intra-Observer ICC

Inter-Observer ICC

ISO Reproducibility

PT

0.978

0.978

1.97°

PI

0.956

0.909

3.85°

SS

0.943

0.877

3.56°

LL

0.935

0.872

5.68°

PI-LL

0.964

0.945

4.94°

TL

0.973

0.971

5.27°

TK

0.903

0.900

6.49°

T9SPI

0.975

0.984

1.04°

T1SPI

0.975

0.995

0.64°

TPA

0.981

0.986

1.97°

T1S

0.717

0.895

4.43°

CL

0.929

0.844

5.87°

TS-CL

0.694

0.799

5.06°

cSVA

0.967

0.914

4.86 mm

SVA

0.978

0.954

4.02 mm

6 7

23 Page 23 of 23