Quasi-automatic 3D reconstruction of the full spine from low-dose biplanar X-rays based on statistical inferences and image analysis

3D external shape analysis and barycentremetry can provide early signs of progression in adolescent idiopathic scoliosis

PURPOSE

Our objective was to analysis the barycentremetry, obtained from the external envelope reconstruction of biplanar radiographs, in adolescent idiopathic scoliosis (AIS) and to determine whether assessing would help predict the distinction between progressive and stable AIS at the early stage.

METHODS

A retrospective study with a multicentre cohort of 205 AIS was conducted. All AIS underwent a biplanar X-ray between 2013 and 2020. Inclusion criteria were Cobb angle between 10° and 25°; Risser sign lower than 3; age higher than 10 years; and no previous treatment. A 3D spine reconstruction was performed, and the barycentremetry parameters were computed, i.e., the center of mass position at the apex and the axial torque at the apex, the upper and lower junction. A severity index, helping to distinguish stable and progressive AIS, was computed on the first radiograph, and weighted according to these parameters. A clinical and radiographic monitoring determined if AIS were classified such a stable or progressive scoliosis.

RESULTS

One hundred and sixty-two AIS were included (i.e., 87 were classified as stable and 75 as progressive). The apex center of mass position was different between the stable and progressive AIS groups (6 mm, SD = 4 mm for the whole cohort; 5 mm, SD = 4 mm for stable AIS versus 7 mm, SD = 4 mm for progressive AIS; p = 0.02). In AIS thoracic, the specificity and positive predictive value of the severity index increased by 19% and 16%, respectively, by adding the apex vertebral axial torque.

CONCLUSION

Early assessment of the external envelope from biplanar X-ray reconstruction of idiopathic scoliosis showed that the apex centre of mass position was significantly different between progressive and stable scoliosis. The inclusion of the axial torque of the apex vertebra in the severity index is promising to help the clinician distinguish between stable and progressive thoracic AIS at an early stage.

LEVEL OF EVIDENCE

II - Prognostic studies.

Arm elevation involves changes in the whole spine: an exploratory study using EOS imaging

BACKGROUND

Few studies have assessed the participation of the spine in arm elevation. The primary aim of this exploratory study was to specify spinal movements during unilateral arm elevation.

METHODS

We used an EOS imaging system to assess 2D global posture (Sagittal Vertical Axis [SVA], T1 and T9 tilt and Central Sacral Line [CSL]) and segmental spine curves (C3-C7 in the sagittal plane only, and T1-T6, T7-T12 and L1-L5 in the sagittal and frontal planes) for four different left arm elevation levels: in the sagittal (Sa) plane (30°Sa: reference position, 140°Sa and 180°Sa), and in the scapular (Sc) plane (180°Sc), in ten right-handed asymptomatic participants (5 women; mean age 24.6 SD 3.0 years]. In addition, we estimated C1, head and pelvic orientation and head and pelvic linear displacement. We used Bayesian statistics (BF10 > 3 indicates a significant variation: moderate, strong, very strong or extreme evidence).

RESULTS

From 140°Sa to 180°Sa or Sc, the significant decrease in SVA and the T1-T9 tilt angles indicated a global backward spine bending (moderate to very strong evidence). The significant reversal of the C3-C7 lordosis at 30°Sa (-1.34 [2.53]°) to kyphosis at 180°Sa (13.88 [3.53]°, strong evidence) and 180°Sc (11.85 [2.75]°, extreme evidence) and the significant decrease in the T7-T12 kyphosis (26.58 [2.84]°at 30°Sa to 16.40 [2.65]° at 180°Sa and 17.60 [2.78]° at 180°Sc [all extreme evidence]) showed a global spine straightening. We found significant pelvic anteversion between 30°Sa and 140°Sa (moderate evidence) and persistent right spine bending and leftward head displacement (extreme evidence). The change in C1 orientation (extreme evidence) showed an atlanto-occipital extension.

CONCLUSION

Simple unconstrained movements of unilateral arm elevation involve the whole spine, pelvis and head, including significant backward spinal bending, a reduction in the low cervical spine lordosis and the thoracic kyphosis, and atlanto-occipital extension.

Automatic 3D reconstruction of vertebrae from orthogonal bi-planar radiographs

When conducting spine-related diagnosis and surgery, the three-dimensional (3D) upright posture of the spine under natural weight bearing is of significant clinical value for physicians to analyze the force on the spine. However, existing medical imaging technologies cannot meet current requirements of medical service. On the one hand, the mainstream 3D volumetric imaging modalities (e.g. CT and MRI) require patients to lie down during the imaging process. On the other hand, the imaging modalities conducted in an upright posture (e.g. radiograph) can only realize 2D projections, which lose the valid information of spinal anatomy and curvature. Developments of deep learning-based 3D reconstruction methods bring potential to overcome the limitations of the existing medical imaging technologies. To deal with the limitations of current medical imaging technologies as is described above, in this paper, we propose a novel deep learning framework, ReVerteR, which can realize automatic 3D Reconstruction of Vertebrae from orthogonal bi-planar Radiographs. With the utilization of self-attention mechanism and specially designed loss function combining Dice, Hausdorff, Focal, and MSE, ReVerteR can alleviate the sample-imbalance problem during the reconstruction process and realize the fusion of the centroid annotation and the focused vertebra. Furthermore, aiming at automatic and customized 3D spinal reconstruction in real-world scenarios, we extend ReVerteR to a clinical deployment-oriented framework, and develop an interactive interface with all functions in the framework integrated so as to enhance human-computer interaction during clinical decision-making. Extensive experiments and visualization conducted on our constructed datasets based on two benchmark datasets of spinal CT, VerSe 2019 and VerSe 2020, demonstrate the effectiveness of our proposed ReVerteR. In this paper, we propose an automatic 3D reconstruction method of vertebrae based on orthogonal bi-planar radiographs. With the 3D upright posture of the spine under natural weight bearing effectively constructed, our proposed method is expected to better support doctors make clinical decision during spine-related diagnosis and surgery.

Research on Three-Dimensional Reconstruction of Ribs Based on Point Cloud Adaptive Smoothing Denoising

The traditional methods for 3D reconstruction mainly involve using image processing techniques or deep learning segmentation models for rib extraction. After post-processing, voxel-based rib reconstruction is achieved. However, these methods suffer from limited reconstruction accuracy and low computational efficiency. To overcome these limitations, this paper proposes a 3D rib reconstruction method based on point cloud adaptive smoothing and denoising. We converted voxel data from CT images to multi-attribute point cloud data. Then, we applied point cloud adaptive smoothing and denoising methods to eliminate noise and non-rib points in the point cloud. Additionally, efficient 3D reconstruction and post-processing techniques were employed to achieve high-accuracy and comprehensive 3D rib reconstruction results. Experimental calculations demonstrated that compared to voxel-based 3D rib reconstruction methods, the 3D rib models generated by the proposed method achieved a 40% improvement in reconstruction accuracy and were twice as efficient as the former.

Spinal axial torque assessment after surgical correction in adolescent idiopathic scoliosis: a new approach to 3D barycentremetry and mass distribution based on biplanar radiographs

PURPOSE

Barycentremetry in adolescent idiopathic scoliosis (AIS) allows the distribution of masses and their loading of the spine to be studied. In particular, the axial torque on the spine has been studied in AIS, but not after surgical correction. Spinal axial torque was studied in AIS before and after surgery.

METHODS

All AIS (Lenke 1 and 3) who underwent posterior spinal fusion surgery at our center in 2019 were included retrospectively. AIS underwent frontal and sagittal biplanar radiographs in the free-standing position before surgery, 4 months after surgery, and at the last follow-up. Their spine and external envelope were reconstructed with validated methods. Spinal axial torque at the apex and the upper and lower end vertebra was calculated. Finally, the preoperative and postoperative values were compared to a previously published reference corridor for asymptomatic subjects.

RESULTS

Twenty-nine patients were included (54 ± 11° Cobb angle, 15 ± 2 years old at surgery). The surgical procedure decreased the Cobb angle by 36° ± 11° and decreased the spinal axial torque at the upper end vertebra by 2.5 N/m (95% CI = [1.9; 3]; p < 0.001), at the apex by 0.6 N/m (95% CI = [0.4; 1]; p = 0.004), at the lower end vertebra by 2 N/m (95% CI = [1.5; 2.8]; p < 0.001). Compared to 95th percentile of torque, which was previously evaluated in asymptomatic subjects, more than 90% of patients had higher values at the upper and lower end vertebrae before surgery. Postoperatively, 62% of patients still had higher torque at the upper end vertebra than asymptomatic subjects, while only 38% patients showed abnormal values at the lower junction.

CONCLUSION

Results of this study confirm that AIS patients show abnormally high spinal axial torque, especially at the end vertebrae, and that this parameter is normalized postoperatively for only a small number of patients.

Evaluation of a method to quantify posture and scapula position using biplanar radiography

BACKGROUND

Recent studies have stated the relevance of having new parameters to quantify the position and orientation of the scapula with patients standing upright. Although biplanar radiography can provide 3D reconstructions of the scapula and the spine, it is not yet possible to acquire these images with patients in the same position.

METHODS

Two pairs of images were acquired, one for the 3D reconstruction of the spine and ribcage and one for the 3D reconstruction of the scapula. Following 3D reconstructions, scapular alignment was performed in two stages, a coarse alignment based on manual annotations of landmarks on the clavicle and pelvis, and an adjusted alignment. Clinical parameters were computed: protraction, internal rotation, tilt and upward rotation. Reproducibility was assessed on an in vivo dataset of upright biplanar radiographs. Accuracy was assessed using supine cadaveric CT-scans and digitally reconstructed radiographs.

FINDINGS

The mean error was less than 2° for all clinical parameters, and the 95 % confidence interval for reproducibility ranged from 2.5° to 5.3°.

INTERPRETATION

The confidence intervals were lower than the variability measured between participants for the clinical parameters assessed, which indicates that this method has the potential to detect different patterns in pathological populations.

Assessment of malalignment at early stage in adolescent idiopathic scoliosis: a longitudinal cohort study

INTRODUCTION

Our objective was to assess abnormalities of the odontoid-hip axis (OD-HA) angle in a mild scoliotic population to determine whether screening for malalignment would help predict the distinction between progressive and stable adolescent idiopathic scoliosis (AIS) at early stage.

MATERIALS AND METHODS

All patients (non-scoliotic and AIS) underwent a biplanar X-ray between 2013 and 2020. In AIS, inclusion criteria were Cobb angle between 10° and 25°; Risser sign lower than 3; age higher than 10 years; and no previous treatment. A 3D spine reconstruction was performed, and the OD-HA was computed automatically. A reference corridor for OD-HA values in non-scoliotic subjects was calculated as the range [5th-95th percentiles]. A severity index, helping to distinguish stable and progressive AIS, was calculated and weighted according to the OD-HA value.

RESULTS

Eighty-three non-scoliotic and 205 AIS were included. The mean coronal and sagittal OD-HA angles in the non-scoliotic group were 0.2° and -2.5°, whereas in AIS values were 0.3° and -0.8°, respectively. For coronal and sagittal OD-HA, 27.5% and 26.8% of AIS were outside the reference corridor compared with 10.8% in non-scoliotic (OR = 3.1 and 3). Adding to the severity index a weighting factor based on coronal OD-HA, for thoracic scoliosis, improved the positive predictive value by 9% and the specificity by 13%.

CONCLUSION

Analysis of OD-HA suggests that AIS patients are almost three times more likely to have malalignment compared with a non-scoliotic population. Furthermore, analysis of coronal OD-HA is promising to help the clinician distinguish between stable and progressive thoracic scoliosis.

Systematic review of imaging comparisons of spinal alignment among standing positions in healthy adolescents or adolescents with idiopathic scoliosis: SOSORT 2023 award winner

PURPOSE

Clinicians detect scoliosis worsening over time using frequent radiographs during growth. Arms must be elevated when capturing sagittal radiographs to visualize the vertebrae, and this may affect the sagittal angles. The aim was to systematically review the published evidence of the effect of arm positions used during radiography on spinal alignment parameters in healthy participants and those with AIS.

METHODS

Design was registered in PROSPERO (CRD42022347494). A search strategy was run in Medline, Embase, CINAHL, and Web of Science. Healthy participants ≥ 10 years old and participants with AIS between 10 and 18 years old, with Cobb angles > 10° were included. Study quality was assessed using the Appraisal tool for Cross-Sectional Studies (AXIS). Meta-analysis was performed where possible.

RESULTS

Overall, 1332 abstracts and 33 full texts were screened. Data was extracted from 7 included studies. The most common positions were habitual standing, fists on clavicle, and active (arms raised unsupported). Kyphosis, lordosis, and sagittal vertical axis (SVA) were most measured. Meta-analysis showed significantly decreased kyphosis (SMD = 0.78, 95%CI 0.48, 1.09) and increased lordosis (SMD = - 1.21, 95%CI - 1.58, - 0.85) when clavicle was compared to standing. Significant posterior shifts in SVA were shown in clavicle compared to standing (MD = 30.59 mm, 95%CI 23.91, 37.27) and active compared to clavicle (MD = - 2.01 mm, 95%CI - 3.38, - 0.64). Cobb angles and rotation were rarely studied (1 study).

CONCLUSION

Meta-analysis evidence showed elevated arm positions modify sagittal measurements compared to standing. Most studies did not report on all relevant parameters. It is unclear which position best represent habitual standing.

3D reconstruction of the scapula from biplanar X-rays for pose estimation and morphological analysis

BACKGROUND

Patient-specific scapular shape in functional posture can be highly relevant to clinical research. Biplanar radiography is a relevant modality for that purpose with already two existing assessment methods. However, they are either time-consuming or lack accuracy. The aim of this study was to propose a new, more user-friendly and accurate method to determine scapular shape.

METHODS

The proposed method relied on simplified manual inputs and an upgraded version of the first 3D estimate based on statistical inferences and Moving-Least Square (MLS) deformation of a template. Then, manual adjustments, with real-time MLS algorithm and contour matching adjustments with an adapted minimal path method, were added to improve the match between the projected 3D model and the radiographic contours. The accuracy and reproducibility of the method were assessed (with 6 and 12 subjects, respectively).

FINDINGS

The shape accuracy was in average under 2 mm (1.3 mm in the glenoid region). The reproducibility study on the clinical parameters found intra-observer 95% confidence intervals under 3 mm or 3° for all parameters, except for glenoid inclination and Critical Shoulder Angle, ranging between 3° and 6°.

INTERPRETATION

This method is a first step towards an accurate reconstruction of the scapula to assess clinical parameters in a functional posture. This can already be used in clinical research on non-pathologic bones to investigate the scapulothoracic joint in functional position.

X-Ray to DRR Images Translation for Efficient Multiple Objects Similarity Measures in Deformable Model 3D/2D Registration

The robustness and accuracy of the intensity-based 3D/2D registration of a 3D model on planar X-ray image(s) is related to the quality of the image correspondences between the digitally reconstructed radiographs (DRR) generated from the 3D models (varying image) and the X-ray images (fixed target). While much effort may be devoted to generating realistic DRR that are similar to real X-rays (using complex X-ray simulation, adding densities information in 3D models, etc.), significant differences still remain between DRR and real X-ray images. Differences such as the presence of adjacent or superimposed soft tissue and bony or foreign structures lead to image matching difficulties and decrease the 3D/2D registration performance. In the proposed method, the X-ray images were converted into DRR images using a GAN-based cross-modality image-to-images translation. With this added prior step of XRAY-to-DRR translation, standard similarity measures become efficient even when using simple and fast DRR projection. For both images to match, they must belong to the same image domain and essentially contain the same kind of information. The XRAY-to-DRR translation also addresses the well-known issue of registering an object in a scene composed of multiple objects by separating the superimposed or/and adjacent objects to avoid mismatching across similar structures. We applied the proposed method to the 3D/2D fine registration of vertebra deformable models to biplanar radiographs of the spine. We showed that the XRAY-to-DRR translation enhances the registration results, by increasing the capture range and decreasing dependence on the similarity measure choice since the multi-modal registration becomes mono-modal.

Sagittal Balance Using Position and Orientation of Each Vertebra in an Asymptomatic Population

STUDY DESIGN

A monocentric, retrospective radiographic study with 99 asymptomatic volunteers.

OBJECTIVE

The authors performed the postural analysis commonly scheduled when evaluating sagittal balance in a vertebra-by-vertebra manner by enrolling an asymptomatic population. They measured the position and angulation of each vertebra to reveal those for which the spatial positioning could be relevant during spinal surgeries.

SUMMARY OF BACKGROUND DATA

Several recent publications detailed the sagittal alignment parameters and focus on global analysis parameters. Some patients with identical commonly evaluated spinal parameters have exhibited very different profiles, with notable differences in vertebral positions and orientations. Therefore, a fine segmental analysis of position of each vertebra could be interest to gain understanding of spine alignment.

MATERIALS AND METHODS

The authors obtained full-spine EOS x-rays of 99 volunteers in the standard free-standing position. We used a validated three-dimensional reconstruction technique to extract current spinal parameters and the positions and angulations of all vertebrae and lumbar discs. Particular attention was paid to the positions and angulations of the apical and transitional vertebrae in the general population and in subgroups according to pelvic incidence (PI).

RESULTS

T1 was the most common transitional cervicothoracic vertebra (in 89.9% of subjects) and was oriented downwards by an average of 22.0° (SD=7.3°, minimum=2.3°, maximum=40.1°). The thoracic apex trio of T5 (22.2%), T6 (28.3%), and T7 (36.4%) were equally found. The transitional thoracolumbar vertebrae were L1 (39.4%) and T12 (33.3%). The lumbar apex was usually the L3-L4 disc (36.4%). T1 seemed to be the transitional vertebra (90%) irrespective of the PI. For the other relevant vertebrae, the greater the PI, the more cranial the vertebra.

CONCLUSIONS

We performed a detailed three-dimensional assessment of overall spinal balance using positional and rotational parameters. The positions and orientations of all vertebrae were specified, particularly the apical and transitional vertebrae.

LEVEL OF EVIDENCE

Level 3.

Institution-wide shape analysis of 3D spinal curvature and global alignment parameters

The spine is an articulated, 3D structure with 6 degrees of translational and rotational freedom. Clinical studies have shown spinal deformities are associated with pain and functional disability in both adult and pediatric populations. Clinical decision making relies on accurate characterization of the spinal deformity and monitoring of its progression over time. However, Cobb angle measurements are time-consuming, are limited by interobserver variability, and represent a simplified 2D view of a 3D structure. Instead, spine deformities can be described by 3D shape parameters, addressing the limitations of current measurement methods. To this end, we develop and validate a deep learning algorithm to automatically extract the vertebral midline (from the upper endplate of S1 to the lower endplate of C7) for frontal and lateral radiographs. Our results demonstrate robust performance across datasets and patient populations. Approximations of 3D spines are reconstructed from the unit normalized midline curves of 20,118 pairs of full spine radiographs belonging to 15,378 patients acquired at our institution between 2008 and 2020. The resulting 3D dataset is used to describe global imbalance parameters in the patient population and to build a statistical shape model to describe global spine shape variations in preoperative deformity patients via eight interpretable shape parameters. The developed method can identify patient subgroups with similar shape characteristics without relying on an existing shape classification system.

Biomechanical Morphing for Personalized Fitting of Scoliotic Torso Skeleton Models

The use of patient-specific biomechanical models offers many opportunities in the treatment of adolescent idiopathic scoliosis, such as the design of personalized braces. The first step in the development of these patient-specific models is to fit the geometry of the torso skeleton to the patient’s anatomy. However, existing methods rely on high-quality imaging data. The exposure to radiation of these methods limits their applicability for regular monitoring of patients. We present a method to fit personalized models of the torso skeleton that takes as input biplanar low-dose radiographs. The method morphs a template to fit annotated points on visible portions of the spine, and it relies on a default biomechanical model of the torso for regularization and robust fitting of hardly visible parts of the torso skeleton, such as the rib cage. The proposed method provides an accurate and robust solution to obtain personalized models of the torso skeleton, which can be adopted as part of regular management of scoliosis patients. We have evaluated the method on ten young patients who participated in our study. We have analyzed and compared clinical metrics on the spine and the full torso skeleton, and we have found that the accuracy of the method is at least comparable to other methods that require more demanding imaging methods, while it offers superior robustness to artifacts such as interpenetration of ribs. Normal-dose X-rays were available for one of the patients, and for the other nine we acquired low-dose X-rays, allowing us to validate that the accuracy of the method persisted under less invasive imaging modalities.

A fresh look at spinal alignment and deformities: Automated analysis of a large database of 9832 biplanar radiographs

We developed and used a deep learning tool to process biplanar radiographs of 9,832 non-surgical patients suffering from spinal deformities, with the aim of reporting the statistical distribution of radiological parameters describing the spinal shape and the correlations and interdependencies between them. An existing tool able to automatically perform a three-dimensional reconstruction of the thoracolumbar spine has been improved and used to analyze a large set of biplanar radiographs of the trunk. For all patients, the following parameters were calculated: spinopelvic parameters; lumbar lordosis; mismatch between pelvic incidence and lumbar lordosis; thoracic kyphosis; maximal coronal Cobb angle; sagittal vertical axis; T1-pelvic angle; maximal vertebral rotation in the transverse plane. The radiological parameters describing the sagittal alignment were found to be highly interrelated with each other, as well as dependent on age, while sex had relatively minor but statistically significant importance. Lumbar lordosis was associated with thoracic kyphosis, pelvic incidence and sagittal vertical axis. The pelvic incidence-lumbar lordosis mismatch was found to be dependent on the pelvic incidence and on age. Scoliosis had a distinct association with the sagittal alignment in adolescent and adult subjects. The deep learning-based tool allowed for the analysis of a large imaging database which would not be reasonably feasible if performed by human operators. The large set of results will be valuable to trigger new research questions in the field of spinal deformities, as well as to challenge the current knowledge.

Anatomy-Aware Inference of the 3D Standing Spine Posture from 2D Radiographs

An important factor for the development of spinal degeneration, pain and the outcome of spinal surgery is known to be the balance of the spine. It must be analyzed in an upright, standing position to ensure physiological loading conditions and visualize load-dependent deformations. Despite the complex 3D shape of the spine, this analysis is currently performed using 2D radiographs, as all frequently used 3D imaging techniques require the patient to be scanned in a prone position. To overcome this limitation, we propose a deep neural network to reconstruct the 3D spinal pose in an upright standing position, loaded naturally. Specifically, we propose a novel neural network architecture, which takes orthogonal 2D radiographs and infers the spine’s 3D posture using vertebral shape priors. In this work, we define vertebral shape priors using an atlas and a spine shape prior, incorporating both into our proposed network architecture. We validate our architecture on digitally reconstructed radiographs, achieving a 3D reconstruction Dice of 0.95, indicating an almost perfect 2D-to-3D domain translation. Validating the reconstruction accuracy of a 3D standing spine on real data is infeasible due to the lack of a valid ground truth. Hence, we design a novel experiment for this purpose, using an orientation invariant distance metric, to evaluate our model’s ability to synthesize full-3D, upright, and patient-specific spine models. We compare the synthesized spine shapes from clinical upright standing radiographs to the same patient’s 3D spinal posture in the prone position from CT.

Personalised gravitational loading of the cervical spine from biplanar X-rays for asymptomatic and clinical subjects in neutral standing position

BACKGROUND

As a leading cause of disability with a high societal and economic cost, it is crucial to better understand risk factors of neck pain and surgical complications. Getting subject-specific external loading is essential for quantifying muscle forces and joint loads but it requires exertion trials and load cells which are uncommon in clinical settings.

METHODS

This paper presents a method to compute the gravitational loading at four levels of the cervical spine (C3C4, C4C5, C5C6, C6C7) in neutral standing position from biplanar radiographs exclusively. The resulting load was decomposed in local disc frames and its components were used to compare different populations: 118 asymptomatic subjects and 46 patients before and after surgery (anterior cervical discectomy and fusion or total disc replacement). Comparisons were performed at C6C7 and the upper level adjacent to surgery.

FINDINGS

Significant changes in gravitational loading were observed with age in healthy subjects as well as in patients after surgery and have been associated with changes in posture.

INTERPRETATION

This approach quantifies the influence of postural changes on gravitational loading on the cervical spine. It represents a simple way to obtain necessary input for muscle force quantification models in clinical routine and to use them for patient evaluation. The study of the subsequent subject-specific spinal loading could help further the understanding of cervical spine biomechanics, degeneration mechanisms and complications following surgery.

Effect of postural alignment alteration with age on vertebral strength

EOS biplane radiographs of 117 subjects between 20 and 83 years were analyzed to compute the upper body lever arm over the L1 vertebra and its impact on vertebral strength. Postural sagittal alignment alteration was observed with age and resulted in a greater lever arm causing vertebral strength to decrease.

PURPOSE

The purpose of this study was to analyze the impact of postural alignment changes with age on vertebral strength using finite element analysis and barycentremetry.

METHODS

A total of 117 subjects from 20 to 83 years were divided in three age groups: young (20 to 40 years, 62 subjects), intermediate (40 to 60 years, 26 subjects), and elderly (60 years and over, 29 subjects). EOS biplane radiographs were acquired, allowing 3D reconstruction of the spine and body envelope as well as spinal, pelvic, and sagittal alignment parameter measurements. A barycentremetry method allowed the estimation of the mass and center of mass (CoM) position of the upper body above L1, relatively to the center of the L1 vertebra (lever arm). To investigate the effect of this lever arm, vertebral strength of a generic finite element model (with constant geometry and mechanical properties for all subjects) was successively computed applying the personalized lever arm of each subject.

RESULTS

A combination of an increase in thoracic kyphosis, cervical lordosis, and pelvic tilt with a loss of lumbar lordosis was observed between the young and the older groups. Sagittal alignment parameters indicated a more forward position as age increased. The lever arm of the CoM above L1 varied from an average of 1 mm backward for the young group, to averages of 10 and 24 mm forward, respectively, for the intermediate and elderly group. As a result, vertebral strength decreased from 2527 N for the young group to 1820 N for the elderly group.

CONCLUSION

The global sagittal alignment modifications observed with age were consistent with the literature. Posture alteration with age reduced vertebral strength significantly in this simplified loading model. Postural alignment seems essential to be considered in the evaluation of osteoporotic patients.

Analysis of apex and transitional vertebra of the spine according to pelvic incidence using orientation and position parameters

OBJECTIVE

To identify the different apex and transitional vertebra according to the shape of the pelvis of individuals despite their difference in sagittal alignment using our measurement system.

METHODS

Full-spine X-rays using EOS in standard stand-position of 99 volunteers were selected (47 women, 52 men, mean age 31 years old). Validated 3D reconstruction technique allows extraction of spinopelvic parameters, and position and rotation of each vertebra and lumbar disks. Subjects were divided into three groups: low PI (lowPI, n = 37), moderate PI (midPI, n = 52), high PI (highPI, n = 10), with, respectively, a PI below 45°, between 45° and 60° and above 60°. Occurrence of specific position and rotation values of apex and transitional vertebra were assessed in each group.

RESULTS

Frequency curves tend to move cranially when the incidence increases except in cervicothoracic where T1 is a constant for all shapes of spine with occurrence approaching 90%. Angulation value of relevant vertebra and lumbar lordosis are significantly positively correlated for the whole population.

CONCLUSIONS

Our study allowed the assessment of the distribution of spine curvatures according to the pelvic incidence. It describes the occurrence of localization of the apex and transitional vertebrae according to pelvic incidence. These results should be taken into account during the analysis of the sagittal balance, especially when planning deformity surgery in adults.

Balance, barycentremetry and external shape analysis in idiopathic scoliosis: What can the physician expect from it?

OBJECTIVE

Our objective was to establish a corridor of normality for the external shape 3D parameters and then to assess these variables in adolescent idiopathic scoliosis (AIS).

METHODS

Adolescent with mild and severe AIS were included prospectively, as well as a control group of asymptomatic subjects. A quasi-automatic 3D reconstruction of the spine and manual 3D reconstruction of the external envelope was performed from biplanar radiography. The center of mass position, the axial intersegmental moment resulting at the apex and junctional vertebrae, and the coronal trunk balance were automatically computed. A normality corridor of asymptomatic subjects was calculated as the range [5^th-95^th percentiles] for external shape parameters at each vertebral level.

RESULTS

Forty-one asymptomatic subjects (19 females; 22 males; 21 yo, SD=4) and sixty AIS (56 females; 4 males; 13 years old, SD=1.9; 30 mild and 30 severe; 34 thoracic curves and 26 thoraco-lumbar or lumbar curves) were included. All parameters based on the external shape showed differences between AIS and controls, as well as between mild and severe scoliosis. For instance, the intersegmental moment applied to the upper junctional vertebra was above the 95^th percentile of controls in 70% of AIS patient. The percentage of severe patients showing parameters higher than the normality corridor was significantly higher than mild patients (p<0.0001).

CONCLUSIONS

The analysis of center of mass, vertebral intersegmental moment and coronal trunk balance parameters appear to be relevant in characterizing the 3D deformity of adolescent idiopathic scoliosis. The upper junctional intersegmental moment seems to be able to distinguish the different stages of curvature severity.

2-step deep learning model for landmarks localization in spine radiographs

In this work we propose to use Deep Learning to automatically calculate the coordinates of the vertebral corners in sagittal x-rays images of the thoracolumbar spine and, from those landmarks, to calculate relevant radiological parameters such as L1–L5 and L1–S1 lordosis and sacral slope. For this purpose, we used 10,193 images annotated with the landmarks coordinates as the ground truth. We realized a model that consists of 2 steps. In step 1, we trained 2 Convolutional Neural Networks to identify each vertebra in the image and calculate the landmarks coordinates respectively. In step 2, we refined the localization using cropped images of a single vertebra as input to another convolutional neural network and we used geometrical transformations to map the corners to the original image. For the localization tasks, we used a differentiable spatial to numerical transform (DSNT) as the top layer. We evaluated the model both qualitatively and quantitatively on a set of 195 test images. The median localization errors relative to the vertebrae dimensions were 1.98% and 1.68% for x and y coordinates respectively. All the predicted angles were highly correlated with the ground truth, despite non-negligible absolute median errors of 1.84°, 2.43° and 1.98° for L1–L5, L1–S1 and SS respectively. Our model is able to calculate with good accuracy the coordinates of the vertebral corners and has a large potential for improving the reliability and repeatability of measurements in clinical tasks.

Effect of curve location on the severity index for adolescent idiopathic scoliosis: a longitudinal cohort study

OBJECTIVES

Adolescent idiopathic scoliosis (AIS) is the most common spinal disorder in children. A severity index was recently proposed to identify the stable from the progressive scoliosis at the first standardized biplanar radiographic exam. The aim of this work was to extend the validation of the severity index and to determine if curve location influences its predictive capabilities.

METHODS

AIS patients with Cobb angle between 10° and 25°, Risser 0-2, and no previous treatment were included. They underwent standing biplanar radiography and 3D reconstruction of the spine and pelvis, which allowed to calculate their severity index. Patients were grouped by curve location (thoracic, thoracolumbar, lumbar). Patients were followed up until skeletal maturity (Risser ≥ 3) or brace prescription. Their outcome was compared to the prediction made by the severity index.

RESULTS

In total, 205 AIS patients were included; 82% of them (155/189, 95% confidence interval [74-90%]) were correctly classified by the index, while 16 patients were unclassified. Positive predictive ratio was 78% and negative predictive ratio was 86%. Specificity (78%) was not significantly affected by curve location, while patients with thoracic and lumbar curves showed higher sensitivity (≥ 89%) than those with thoracolumbar curves (74%).

CONCLUSIONS

In this multicentric cohort of 205 patients, the severity index was used to predict the risk of progression from mild to moderate scoliosis, with similar results of typical major curve types. This index represents a novel tool to aid the clinician and the patient in the modulation of the follow-up and, for progressive patients, their decision for brace treatment.

KEY POINTS

• The severity index of adolescent idiopathic scoliosis has the potential to detect patients with progressive scoliosis as early as the first exam. • Out of 205 patients, 82% were correctly classified as either stable or progressive by the severity index. • The location of the main curve had small effect on the predictive capability of the index.

Toward Automated 3D Spine Reconstruction from Biplanar Radiographs Using CNN for Statistical Spine Model Fitting

To date, 3D spine reconstruction from biplanar radiographs involves intensive user supervision and semi-automated methods that are time-consuming and not effective in clinical routine. This paper proposes a new, fast, and automated 3D spine reconstruction method through which a realistic statistical shape model of the spine is fitted to images using convolutional neural networks (CNN). The CNNs automatically detect the anatomical landmarks controlling the spine model deformation through a hierarchical and gradual iterative process. The performance assessment used a set of 68 biplanar radiographs, composed of both asymptomatic subjects and adolescent idiopathic scoliosis patients, in order to compare automated reconstructions with ground truths build using multiple experts-supervised reconstructions. The mean (SD) errors of landmark locations (3D Euclidean distances) were 1.6 (1.3) mm, 1.8 (1.3) mm, and 2.3 (1.4) mm for the vertebral body center, endplate centers, and pedicle centers, respectively. The clinical parameters extracted from the automated 3D reconstruction (reconstruction time is less than one minute) presented an absolute mean error between 2.8° and 4.7° for the main spinal parameters and between 1° and 2.1° for pelvic parameters. Automated and expert's agreement analysis reported that, on average, 89% of automated measurements were inside the expert's confidence intervals. The proposed automated 3D spine reconstruction method provides an important step that should help the dissemination and adoption of 3D measurements in clinical routine.

Quasi-automatic early detection of progressive idiopathic scoliosis from biplanar radiography: a preliminary validation

PURPOSE

To validate the predictive power and reliability of a novel quasi-automatic method to calculate the severity index of adolescent idiopathic scoliosis (AIS).

METHODS

Fifty-five AIS patients were prospectively included (age 10-15, Cobb 16° ± 4°). Patients underwent low-dose biplanar X-rays, and a novel fast method for 3D reconstruction of the spine was performed. They were followed until skeletal maturity (stable patients) or brace prescription (progressive patients). The severity index was calculated at the first examination, based on 3D parameters of the scoliotic curve, and it was compared with the patient's final outcome (progressive or stable). Three operators have repeated the 3D reconstruction twice for a subset of 30 patients to assess reproducibility (through Cohen's kappa and intra-class correlation coefficient).

RESULTS

Eighty-five percentage of the patients were correctly classified as stable or progressive by the severity index, with a sensitivity of 92% and specificity of 74%. Substantial intra-operator agreement and good inter-operator agreement were observed, with 80% of the progressive patients correctly detected at the first examination. The novel severity index assessment took less than 4 min of operator time.

CONCLUSIONS

The fast and semiautomatic method for 3D reconstruction developed in this work allowed for a fast and reliable calculation of the severity index. The method is fast and user friendly. Once extensively validated, this severity index could allow very early initiation of conservative treatment for progressive patients, thus increasing treatment efficacy and therefore reducing the need for corrective surgery. These slides can be retrieved under Electronic Supplementary Material.