H. Labelle

Assessment of the 3-D reconstruction and high-resolution geometrical modeling of the human skeletal trunk from 2-D radiographic images

This paper presents an in vivo validation of a method for the three-dimensional (3-D) high-resolution modeling of the human spine, rib cage, and pelvis for the study of spinal deformities. The method uses an adaptation of a standard close-range photogrammetry method called direct linear transformation to reconstruct the 3-D coordinates of anatomical landmarks from three radiographic images of the subjects trunk. It then deforms in 3-D 1-mm-resolution anatomical primitives (reference bones) obtained by serial computed tomography-scan reconstruction of a dry specimen. The free-form deformation is calculated using dual kriging equations. In vivo validation of this method on 40 scoliotic vertebrae gives an overall accuracy of 3.3 /spl plusmn/ 3.8 mm, making it an adequate tool for clinical studies and mechanical analysis purposes.

Morphometric evaluations of personalised 3D reconstructions and geometric models of the human spine

In the past, several techniques have been developed to study and analyse the 3D characteristics of the human spine: multi-view radiographic or biplanar 3D reconstructions, CT-scan 3D reconstructions and geometric models. Extensive evaluations of three of these techniques that are routinely used at Sainte-Justine Hospital (Montréal, Canada) are presented. The accuracy of these methods is assessed by comparing them with precise measurements made with a coordinate measuring machine on 17 thoracic and lumbar vertebrae (T1-L5) extracted from a normal cadaveric spine specimen. Multi-view radiographic 3D reconstructions are evaluated for different combinations of X-ray views: lateral (LAT), postero-anterior with normal incidence (PAOo) and postero-anterior with 20o angled down incidence (PA20o). The following accuracies are found for these reconstructions obtained from different radiographic setups: 2.1±1.5 mm for the combination with PAOo-LAT views, and 5.6±4.5 mm for the PAOo-PA20o stereopair. Higher errors are found in the postero-anterior direction, especially for the PAOo-PA20o view combination. Pedicles are found to be the most precise landmarks. Accuracy for CT-scan 3D reconstructions is about 1.1±0.8 mm. As for a geometric model built using a multiview radiographic reconstruction based on six landmarks per vertebra, accuracies of about 2.6±2.4 mm for landmarks and 2.3±2.0 mm for morphometric parameters are found. The geometric model and 3D reconstruction techniques give accurate information, at low X-ray dose. The accuracy assessment of the techniques used to study the 3D characteristics of the human spine is important, because it allows better and more efficient quantitative evaluations of spinal dysfunctions and their treatments, as well as biomechanical modelling of the spine.

A hierarchical statistical modeling approach for the unsupervised 3-D biplanar reconstruction of the scoliotic spine

This paper presents a new and accurate three-dimensional (3-D) reconstruction technique for the scoliotic spine from a pair of planar and conventional (postero-anterior with normal incidence and lateral) calibrated radiographic images. The proposed model uses a priori hierarchical global knowledge, both on the geometric structure of the whole spine and of each vertebra. More precisely, it relies on the specification of two 3-D statistical templates. The first, a rough geometric template on which rigid admissible deformations are defined, is used to ensure a crude registration of the whole spine. An accurate 3-D reconstruction is then performed for each vertebra by a second template on which nonlinear admissible global, as well as local deformations, are defined. Global deformations are modeled using a statistical modal analysis of the pathological deformations observed on a representative scoliotic vertebra population. Local deformations are represented by a first-order Markov process. This unsupervised coarse-to-fine 3-D reconstruction procedure leads to two separate minimization procedures efficiently solved in our application with evolutionary stochastic optimization algorithms. In this context, we compare the results obtained with a classical genetic algorithm (GA) and a recent Exploration Selection (ES) technique. This latter optimization method with the proposed 3-D reconstruction model, is tested on several pairs of biplanar radiographic images with scoliotic deformities. The experiments reported in this paper demonstrate that the discussed method is comparable in terms of accuracy with the classical computed-tomography-scan technique while being unsupervised and while requiring only two radiographic images and a lower amount of radiation for the patient.

Optimization method for 3D bracing correction of scoliosis using a finite element model

Abstract Scoliosis is a complex three-dimensional deformity of the spine and rib cage frequently treated by brace. Although bracing produces significant correction in the frontal plane, it generally reduces the normal sagittal plane curvatures and has limited effect in the transverse plane. The goal of this study is to develop a new optimization approach using a finite element model of the spine and rib cage in order to find optimal correction patterns. The objective function to be minimized took account of coronal and sagittal offsets from a normal spine at the thoracic and lumbar apices as well as the rib hump. Two different optimization studies were performed using the finite element model, which was personalized to the geometry of 20 different scoliotic patients. The first study took into account only the thoracic deformity, while the second considered both the thoracic and lumbar deformities. The optimization produced an average of 56% and 51% reduction of the objective function respectively in the two studies. Optimal forces were mostly located on the convex side of the curve. This study demonstrates the feasibility of using an optimization approach with a finite element model of the trunk to analyze the biomechanics of bracing, and may be useful in the design of new and more effective braces.

Preoperative Planning Simulator for Spinal Deformity Surgeries

Study Design. Proof of concept of a spine surgery simulator (S3) for the assessment of scoliosis instrumentation configuration strategies. Objective. To develop and assess a surgeon-friendly spine surgery simulator that predicts the correction of a scoliotic spine as a function of the patient characteristics and instrumentation variables. Summary of Background Data. There is currently no clinical tool sufficiently user-friendly, reliable and refined for the preoperative planning and prediction of correction using different instrumentation configurations. Methods. A kinetic model using flexible mechanisms has been developed to represent patient-specific spine geometry and flexibility, and to simulate individual substeps of correction with an instrumentation system. The surgeon-friendly simulator interface allows interactive specification of the instrumentation components, surgical correction maneuvers and display of simulation results. Results. The simulations of spinal instrumentation procedures of 10 scoliotic cases agreed well with postoperative results and the expected behavior of the instrumented spine (average Cobb angle differences of 3.5° to 4.6° in the frontal plane and of 3.6° to 4.7° in the sagittal plane). Forces generated at the implant-vertebra link were mostly below reported pull-out values, with more important values at the extremities of the instrumentation. Conclusion. The spine surgery simulator S3 has proven its technical feasibility and clinical relevance to assist in the preoperative planning of instrumentation strategies for the correction of scoliotic deformities.

Progression of vertebral and spinal three-dimensional deformities in adolescent idiopathic scoliosis : A longitudinal study

Study Design. The evolution of scoliotic descriptors was analyzed from three-dimensionally reconstructed spines and assessed statistically in a group of adolescents with progressive idiopathic scoliosis. Objectives. To conduct an intrasubject longitudinal study quantifying evolution of two- and three-dimensional geometrical descriptors characterizing the scoliotic spine and vertebral deformities. Summary of Background Data. The data available on geometric descriptors usually are based on cross-sectional studies comparing scoliotic configurations of different individuals. The literature reports very few longitudinal studies that evaluated different phases of scoliotic progression in the same patients. Methods. The evolution of regional and local descriptors between two scoliotic visits was analyzed in 28 adolescents with scoliosis. Several statistical analyses were performed to determine how spinal curvatures and vertebral deformities change during scoliosis progression. Results. At the thoracic level, vertebral wedging increases with curve severity in a relatively consistent pattern for most patients with scoliosis. Axial rotation mainly increases toward curve convexity with scoliosis severity, worsening the progression of vertebral body deformities. No consistent evolution is associated with the angular orientation of the maximum wedging. Thoracic kyphosis varies considerably among subjects. Both increasing and decreasing kyphosis are observed in nonnegligible proportions. A decrease in kyphosis is associated with a shift in the plane of maximum deformity toward the frontal plane, which worsens the three-dimensional shape of the spine. Conclusions. The results of this study challenge the existence of a typical scoliotic evolution pattern and suggest that scoliotic evolution is quite variable and patient specific.

Simulation of progressive deformities in adolescent idiopathic scoliosis using a biomechanical model integrating vertebral growth modulation

While the etiology and pathogenesis of adolescent idiopathic scoliosis are still not well understood, it is generally recognized that it progresses within a biomechanical process involving asymmetrical loading of the spine and vertebral growth modulation. This study intends to develop a finite element model incorporating vertebral growth and growth modulation in order to represent the progression of scoliotic deformities. The biomechanical model was based on experimental and clinical observations, and was formulated with variables integrating a biomechanical stimulus of growth modulation along directions perpendicular (x) and parallel (y, z) to the growth plates, a sensitivity factor beta to that stimulus and time. It was integrated into a finite element model of the thoracic and lumbar spine, which was personalized to the geometry of a female subject without spinal deformity. An imbalance of 2 mm in the right direction at the 8th thoracic vertebra was imposed and two simulations were performed: one with only growth modulation perpendicular to growth plates (Sim1), and the other one with additional components in the transverse plane (Sim2). Semi-quantitative characterization of the scoliotic deformities at each growth cycle was made using regional scoliotic descriptors (thoracic Cobb angle and kyphosis) and local scoliotic descriptors (wedging angle and axial rotation of the thoracic apical vertebra). In all simulations, spinal profiles corresponded to clinically observable configurations. The Cobb angle increased non-linearly from 0.3 degree to 34 degrees (Sim1) and 20 degrees (Sim2) from the first to last growth cycle, adequately reproducing the amplifying thoracic scoliotic curve. The sagittal thoracic profile (kyphosis) remained quite constant. Similarly to clinical and experimental observations, vertebral wedging angle of the thoracic apex progressed from 2.6 degrees to 10.7 degrees (Sim1) and 7.8 degrees (Sim2) with curve progression. Concomitantly, vertebral rotation of the thoracic apex increased of 10 degrees (Sim1) and 6 degrees (Sim2) clockwise, adequately reproducing the evolution of axial rotation reported in several studies. Similar trends but of lesser magnitude (Sim2) suggests that growth modulation parallel to growth plates tend to counteract the growth modulation effects in longitudinal direction. Overall, the developed model adequately represents the self-sustaining progression of vertebral and spinal scoliotic deformities. This study demonstrates the feasibility of the modeling approach, and compared to other biomechanical studies of scoliosis it achieves a more complete representation of the scoliotic spine.

Biomechanical modelling of orthotic treatment of the scoliotic spine including a detailed representation of the brace-torso interface

As part of the development of new modelling tools for the simulation and design of brace treatment of scoliosis, a finite element model of a brace and its interface with the torso was proposed. The model was adapted to represent one scoliotic adolescent girl treated with a Boston brace. The 3D geometry was acquired using multiview radiographs. The model included the osseo-ligamentous structures, thoracic and abdominal soft tissues, brace foam and shell, and brace-torso interface. The simulations consisted of brace opening to include the patients trunk followed by brace closing. To validate the model, the resulting geometry was compared with the real in-brace geometry, and the resulting contact reaction forces at the brace-torso interface were compared with the equivalent forces calculated from pressure measurements made on the in-brace patient. Differences between coronal equivalent and reaction forces were less than 7N. However, sagittal reaction forces (47N) were computed on the abdomen, whereas negligible equivalent forces were measured. The simulated geometry presented partially reduced coronal Cobb angles (1–40), over-corrected sagittal Cobb angles and maximum deformation plane (50), completely corrected coronal shift, and sagittal shift and rib humps that were not corrected. This study demonstrated the feasibility of a new approach that represents the load transfer from the brace to the spine more realistically than does the direct application of forces.

Three-dimensional measurement of wedged scoliotic vertebrae and intervertebral disks.

Abstract Idiopathic scoliosis involves complex spinal intrinsic deformations such as the wedging of vertebral bodies (VB) and intervertebral disks (ID), and it is obvious that the clinical evaluation obtained by the spinal projections on the two-dimensional (2D) radiographic planes do not give a full and accurate interpretation of scoliotic deformities. This paper presents a method that allows reconstruction in 3D of the vertebral body endplates and measurement of the 3D wedging angles. This approach was also used to verify whether 2D radiographic measurements could lead to a biased evaluation of scoliotic spine wedging. The 3D reconstruction of VB contours was done using calibrated biplanar X-rays and an iterative projection computer procedure that fits 3D oriented ellipses of adequate diameters onto the 3D endplate contours. “3D wedging angles” of the VB and ID (representing the maximum angle between adjacent vertebrae) as well as their angular locations with respect to the vertebral frontal planes were computed by finding the positions of the shortest and longest distances between consecutive endplates along their contour. This method was extensively validated using several approaches: (1) by comparing the 3D reconstructed endplates of a cadaveric functional unit (T8-T9) with precise 3D measurements obtained using a coordinate measuring machine for 11 different combinations of vertebral angular positions; (2) by a sensitivity study on 400 different vertebral segments mathematically generated, with errors randomly introduced on the digitized points (standard deviations of 0.5, 1, 2, and 3 mm); (3) by comparing the clinical wedging measurements (on postero-anterior and lateral radiographs) at the thoracic apical level of 34 scoliotic patients (15° < Cobb < 45°) to the computed values. Mean errors for the 11 vertebral positions were 0.5 ± 0.4 mm for VB thickness, less than 2.2° for endplate orientation, and about 11° (3 mm) for the location of the maximum 3D wedging angle along the endplate contour. The errors below 2 mm (introduced on the digitized points) slightly affected the 3D wedging angle (< 2°) and its location (< 4°) for the ID. As for the clinical evaluation, average angular errors were less than 0.4° in the radiographic frontal and lateral planes. The mean 3D wedged angles were about 4.9°± 1.9° for the VB and 6.0°± 1.7° for the ID. Linear relations were found between the 2D and the 3D angles, but the 3D angles were located on diagonal planes statistically different than the radiographic ones (between 100° and 221°). There was no statistical relation between the 2D radiographic angles and the locations of the 3D intervertebral wedging angles. These results clearly indicate that VB and ID endplates are wedged in 3D, and that measurements on plain radiographs allow incomplete evaluation of spinal wedging. Clinicians should be aware of these limitations while using wedging measurements from plain radiographs for diagnosis and/or research on scoliotic deformities.

Long-term three-dimensional changes of the spine after posterior spinal instrumentation and fusion in adolescent idiopathic scoliosis.

Abstract This is a prospective study comparing the short- and long-term three-dimensional (3D) changes in shape, length and balance of the spine after spinal instrumentation and fusion in a group of adolescents with idiopathic scoliosis. The objective of the study was to evaluate the stability over time of the postoperative changes of the spine after instrumentation with multi rod, hook and screw instrumentation systems. Thirty adolescents (average age: 14.5 ± 1.6 years) undergoing surgery by a posterior approach had computerized 3D reconstructions of the spine done at an average of 3 days preoperatively (stage I), and 2 months (stage II) and 2,5 years (stage III) after surgery, using a digital multi-planar radiographic technique. Stages I, II and III were compared using various geometrical parameters of spinal length, curve severity, and orientation. Significant improvement of curve magnitude between stages I and II was documented in the frontal plane for thoracic and lumbar curves, as well as in the orientation of the plane of maximum deformity, which was significantly shifted towards the sagittal plane in thoracic curves. However, there was a significant loss of this correction between stages II and III. Slight changes were noted in apical vertebral rotation, in thoracic kyphosis and in lumbar lordosis. Spinal length and height were significantly increased at stage II, but at long-term follow-up spinal length continued to increase while spinal height remained similar. These results indicate that although a significant 3D correction can be obtained after posterior instrumentation and fusion, a significant loss of correction and an increase in spinal length occur in the years following surgery, suggesting that a crankshaft phenomenon may be an important factor altering the long-term 3D correction after posterior instrumentation of the spine for idiopathic scoliosis.