Evaluation of automated statistical shape model based knee kinematics from biplane fluoroscopy

Scan-Free and Fully Automatic Tracking of Native Knee Anatomy from Dynamic Stereo-Radiography with Statistical Shape and Intensity Models

Kinematic tracking of native anatomy from stereo-radiography provides a quantitative basis for evaluating human movement. Conventional tracking procedures require significant manual effort and call for acquisition and annotation of subject-specific volumetric medical images. The current work introduces a framework for fully automatic tracking of native knee anatomy from dynamic stereo-radiography which forgoes reliance on volumetric scans. The method consists of three computational steps. First, captured radiographs are annotated with segmentation maps and anatomic landmarks using a convolutional neural network. Next, a non-convex polynomial optimization problem formulated from annotated landmarks is solved to acquire preliminary anatomy and pose estimates. Finally, a global optimization routine is performed for concurrent refinement of anatomy and pose. An objective function is maximized which quantifies similarities between masked radiographs and digitally reconstructed radiographs produced from statistical shape and intensity models. The proposed framework was evaluated against manually tracked trials comprising dynamic activities, and additional frames capturing a static knee phantom. Experiments revealed anatomic surface errors routinely below 1.0 mm in both evaluation cohorts. Median absolute errors of individual bone pose estimates were below 1.0∘ or mm for 15 out of 18 degrees of freedom in both evaluation cohorts. Results indicate that accurate pose estimation of native anatomy from stereo-radiography may be performed with significantly reduced manual effort, and without reliance on volumetric scans.

Fully automatic tracking of native glenohumeral kinematics from stereo-radiography

The current work introduces a system for fully automatic tracking of native glenohumeral kinematics in stereo-radiography sequences. The proposed method first applies convolutional neural networks to obtain segmentation and semantic key point predictions in biplanar radiograph frames. Preliminary bone pose estimates are computed by solving a non-convex optimization problem with semidefinite relaxations to register digitized bone landmarks to semantic key points. Initial poses are then refined by registering computed tomography-based digitally reconstructed radiographs to captured scenes, which are masked by segmentation maps to isolate the shoulder joint. A particular neural net architecture which exploits subject-specific geometry is also introduced to improve segmentation predictions and increase robustness of subsequent pose estimates. The method is evaluated by comparing predicted glenohumeral kinematics to manually tracked values from 17 trials capturing 4 dynamic activities. Median orientation differences between predicted and ground truth poses were 1.7∘ and 8.6∘ for the scapula and humerus, respectively. Joint-level kinematics differences were less than 2∘ in 65%, 13%, and 63% of frames for XYZ orientation DoFs based on Euler angle decompositions. Automation of kinematic tracking can increase scalability of tracking workflows in research, clinical, or surgical applications.

Integration of statistical shape modeling and alternating interpolation-based model tracking technique for measuring knee kinematics in vivo using clinical interleaved bi-plane fluoroscopy

Background

A 2D fluoroscopy/3D model-based registration with statistical shape modeling (SSM)-reconstructed subject-specific bone models will help reduce radiation exposure for 3D kinematic measurements of the knee using clinical alternating bi-plane fluoroscopy systems. The current study aimed to develop such an approach and evaluate in vivo its accuracy and identify the effects of the accuracy of SSM models on the kinematic measurements.

Methods

An alternating interpolation-based model tracking (AIMT) approach with SSM-reconstructed subject-specific bone models was used for measuring 3D knee kinematics from dynamic alternating bi-plane fluoroscopy images. A two-phase optimization scheme was used to reconstruct subject-specific knee models from a CT-based SSM database of 60 knees using one, two, or three pairs of fluoroscopy images. Using the CT-reconstructed model as a benchmark, the performance of the AIMT with SSM-reconstructed models in measuring bone and joint kinematics during dynamic activity was evaluated in terms of mean target registration errors (mmTRE) for registered bone poses and the mean absolute differences (MAD) for each motion component of the joint poses.

Results

The mmTRE of the femur and tibia for one image pair were significantly greater than those for two and three image pairs without significant differences between two and three image pairs. The MAD was 1.16 to 1.22° for rotations and 1.18 to 1.22 mm for translations using one image pair. The corresponding values for two and three image pairs were 0.75 to 0.89° and 0.75 to 0.79 mm; and 0.57 to 0.79° and 0.6 to 0.69 mm, respectively. The MAD values for one image pair were significantly greater than those for two and three image pairs without significant differences between two and three image pairs.

Conclusions

An AIMT approach with SSM-reconstructed models was developed, enabling the registration of interleaved fluoroscopy images and SSM-reconstructed models from more than one asynchronous fluoroscopy image pair. This new approach had sub-millimeter and sub-degree measurement accuracy when using more than one image pair, comparable to the accuracy of CT-based methods. This approach will be helpful for future kinematic measurements of the knee with reduced radiation exposure using 3D fluoroscopy with clinically alternating bi-plane fluoroscopy systems.

The accuracy of statistical shape models in predicting bone shape: A systematic review

BACKGROUND

This systematic review aims to ascertain how accurately 3D models can be predicted from two-dimensional (2D) imaging utilising statistical shape modelling.

METHODS

A systematic search of published literature was conducted in September 2022. All papers which assessed the accuracy of 3D models predicted from 2D imaging utilising statistical shape models and which validated the models against the ground truth were eligible.

RESULTS

2127 papers were screened and a total of 34 studies were included for final data extraction. The best overall achievable accuracy was 0.45 mm (root mean square error) and 0.16 mm (average error).

CONCLUSION

Statistical shape modelling can predict detailed 3D anatomical models from minimal 2D imaging. Future studies should report the intended application domain of the model, the level of accuracy required, the underlying demographics of subjects, and the method in which accuracy was calculated, with root mean square error recommended if appropriate.

A shape-partitioned statistical shape model for highly deformed femurs using X-ray images

To develop a patient-specific 3 D reconstruction of a femur modeled using the statistical shape model (SSM) and X-ray images, it is assumed that the target shape is not outside the range of variations allowed by the SSM built from a training dataset. We propose the shape-partitioned statistical shape model (SPSSM) to cover significant variations in the target shape. This model can divide a shape into several segments of anatomical interest. We break up the eigenvector matrix into the corresponding representative matrices for the SPSSM by preserving the relevant rows of the original matrix without segmenting the shape and building an independent SSM for each segment. To quantify the reconstruction error of the proposed method, we generated two groups of deformation models of the femur which cannot be easily represented by the conventional SSM. One group of femurs had an anteversion angle deformation, and the other group of femurs had two different scales of the femoral head. Each experiment was performed using the leave-one-out method for twelve femurs. When the femoral head was rotated by 30°, the average reconstruction error of the conventional SSM was 5.34 mm, which was reduced to 3.82 mm for the proposed SPSSM. When the femoral head size was decreased by 20%, the average reconstruction error of the SSM was 4.70 mm, which was reduced to 3.56 mm for the SPSSM. When the femoral head size was increased by 20%, the average reconstruction error of the SSM was 4.28 mm, which was reduced to 3.10 mm for the SPSSM. The experimental results for the two groups of deformation models showed that the proposed SPSSM outperformed the conventional SSM.

A parametric head geometry model accounting for variation among adolescent and young adult populations

BACKGROUND AND OBJECTIVE

Modeling the size and shape of human skull and scalp is essential for head injury assessment, design of helmets and head-borne equipment, and many other safety applications. Finite element (FE) head models are important tools to assess injury risks and design personal protective equipment. However, current FE head models are mainly developed based on the midsize male, failing to account for the significant morphological variation that exists in the skull and brain. The objective of this study was to develop a statistical head geometry model that accounts for size and shape variations among the adolescent and young adult population.

METHODS

To represent subject-specific geometry using a homologous mesh, threshold-based segmentation of head CT scans of 101 subjects between 14 and 25 years of age was performed, followed by landmarking, mesh morphing, and projection. Skull and scalp statistical geometry models were then developed as functions of age, sex, stature, BMI, head length, head breadth, and tragion-to-top of head using generalized Procrustes analysis (GPA), principal component analysis (PCA) and multivariate regression analysis.

RESULTS

The statistical geometry models account for a high percentage of morphological variations in scalp geometry (R²=0.63), outer skull geometry (R²=0.66), inner skull geometry (R²=0.55), and skull thickness (error < 1 mm) CONCLUSIONS: Skull and scalp statistical geometry models accounts for size and shape variations among the adolescent and young adult population were developed as functions of subject covariates. These models may serve as the geometric basis to develop individualized head FE models for injury assessment and design of head-borne equipment.

Three-Dimensional Subject-Specific Knee Shape Reconstruction with Asynchronous Fluoroscopy Images Using Statistical Shape Modeling

Background and objectives: Statistical shape modeling (SSM) based on computerized tomography (CT) datasets has enabled reasonably accurate reconstructions of subject-specific 3D bone morphology from one or two synchronous radiographs for clinical applications. Increasing the number of radiographic images may increase the reconstruction accuracy, but errors related to the temporal and spatial asynchronization of clinical alternating bi-plane fluoroscopy may also increase. The current study aimed to develop a new approach for subject-specific 3D knee shape reconstruction from multiple asynchronous fluoroscopy images from 2, 4, and 6 X-ray detector views using a CT-based SSM model; and to determine the optimum number of planar images for best accuracy via computer simulations and in vivo experiments.

Methods: A CT-based SSM model of the knee was established from 60 training models in a healthy young Chinese male population. A new two-phase optimization approach for 3D subject-specific model reconstruction from multiple asynchronous clinical fluoroscopy images using the SSM was developed, and its performance was evaluated via computer simulation and in vivo experiments using one, two and three image pairs from an alternating bi-plane fluoroscope.

Results: The computer simulation showed that subject-specific 3D shape reconstruction using three image pairs had the best accuracy with RMSE of 0.52 ± 0.09 and 0.63 ± 0.085 mm for the femur and tibia, respectively. The corresponding values for the in vivo study were 0.64 ± 0.084 and 0.69 ± 0.069 mm, respectively, which was significantly better than those using one image pair (0.81 ± 0.126 and 0.83 ± 0.108 mm). No significant differences existed between using two and three image pairs.

Conclusion: A new two-phase optimization approach was developed for SSM-based 3D subject-specific knee model reconstructions using more than one asynchronous fluoroscopy image pair from widely available alternating bi-plane fluoroscopy systems in clinical settings. A CT-based SSM model of the knee was also developed for a healthy young Chinese male population. The new approach was found to have high mode reconstruction accuracy, and those for both two and three image pairs were much better than for a single image pair. Thus, two image pairs may be used when considering computational costs and radiation dosage. The new approach will be useful for generating patient-specific knee models for clinical applications using multiple asynchronous images from alternating bi-plane fluoroscopy widely available in clinical settings. The current SSM model will serve as a basis for further inclusion of training models with a wider range of sizes and morphological features for broader applications.

Automatic tracking of healthy joint kinematics from stereo-radiography sequences

Kinematic tracking of healthy joints in radiography sequences is frequently performed by maximizing similarities between computed perspective projections of 3D computer models and corresponding objects' appearances in radiographic images. Significant human effort associated with manual tracking presents a major bottleneck in biomechanics research methods and limits the scale of target applications. The current work introduces a method for fully-automatic tracking of tibiofemoral and patellofemoral kinematics in stereo-radiography sequences for subjects performing dynamic activities. The proposed method involves the application of convolutional neural networks for annotating radiographs and a multi-stage optimization pipeline for estimating bone pose based on information provided by neural net predictions. Predicted kinematics are evaluated by comparing against manually-tracked trends across 20 distinct trials. Median absolute differences below 1.5 millimeters or degrees for 6 tibiofemoral and 3 patellofemoral degrees of freedom demonstrate the utility of our approach, which improves upon previous semi-automatic methods by enabling end-to-end automation. Implementation of a fully-automatic pipeline for kinematic tracking will benefit evaluation of human movement by enabling large-scale studies of healthy knee kinematics.

A patient-specific 3D model of the knee to compare the femoral rollback before and after total knee arthroplasty (TKA)

Purpose

Total knee arthroplasty (TKA) is nowadays performed as a standard procedure on a large number of patients suffering from arthrosis. Replacing the knee joint causes changes in the geometry and kinematics of the knee, which are unique to each individual. This research focuses on the method to detect these changes after TKA and on the impact on the knee movement. This approach could reduce complications in patients with post-operative pain and reduce the number of revisions.

Methods

A 3D model of a patient’s knee was made by measuring the movement with a medically certified infrared stereo camera. This measurement was combined with the 3D model of the patient’s bones, previously segmented from the CT scan. This model is printed in 3D, one part being the mechanism that follows the movement of the patient, and the other part being the 3D copy of the femur and tibia bones. The knee replacement operation is performed directly on the model and the resulting rollback is being measured before and after TKA.

Results

We observe a difference in the rollback before and after TKA on the 3D printed model. The variation in size and shape of the femoral implant compared to the natural femur condyles is one of the reasons for the changes in the rollback effect. The rollback is half as large after the prosthesis insertion, which confirms the fact that the femoral prosthesis geometry influences the knee kinematics.

Conclusions

In this study, a first 3D model combining the patient-specific kinematic and the geometry of his bones has been constructed. This model allows the surgeon to validate the plan of the operation, but also to understand the problems and consequences generated by the prosthesis insertion. The rollback is one of the most important motion of the knee joint and this behavior could be quantified, providing comparative analysis of the knee joint before and after the operation. As a future study, the model could be used to analyse more parameters of the TKA such as the impact of different implantation methods.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40634-020-00319-6.

Patient-specific instrumentation improves functional kinematics of minimally-invasive total knee replacements as revealed by computerized 3D fluoroscopy

BACKGROUND AND OBJECTIVES

Minimally-invasive total knee arthroplasty (MIS-TKA) has demonstrated very good short-term success, but its mid- to long-term results remain inconclusive. The success may be related to the tradeoff between a small incision and accurate positioning of the implant components. Patient-specific instrumentation (PSI) aims to improve the accuracy in restoring the knee axis and the clinical outcomes for MIS-TKA, but the results are yet to be confirmed by accurate assessment during functional activities. The purpose of the current study was to measure and compare the in vivo three-dimensional (3D) rigid-body and surface kinematics of MIS-TKA implanted with and without PSI during isolated knee active flexion/extension and sit-to-stand using state-of-the-art 3D model-based fluoroscopy technology.

METHODS

Ten patients treated for advanced medial knee osteoarthritis by MIS-TKA without PSI (non-PSI group) and nine with PSI (PSI group) participated in the current study. Each subject performed non-weight-bearing knee flexion/extension and sit-to-stand tasks while the motion of the prosthetic knee was under bi-plane fluoroscopy surveillance. The computer models of each of the knee prosthesis components were registered to the measured fluoroscopy images for each time frame via a novel validated 3D fluoroscopy method. Non-parametric 1-tailed Mann-Whitney tests were performed to detect the differences in the joint and surface kinematic variables every 10° of knee flexion between the non-PSI and PSI groups. The 1-tailed significance level was at α = 0.05.

RESULTS

The PSI group showed clear, coupled flexion/internal rotation during activities, while the non-PSI group remained roughly at an externally rotated position with slight internal rotations. The coupled rotation in the PSI group was accompanied by an anterior displacement of the medial contact and a posterior displacement of the lateral contact, which was different from the screw-home mechanism. Neither of the two groups showed the normal roll-back phenomenon, i.e., posterior translation of the femur relative to the tibia during knee flexion.

CONCLUSIONS

With the state-of-the-art 3D fluoroscopy method, differences in both the rigid-body and surface kinematics of the prosthetic knees between MIS-TKA with and without PSI were identified. Patients with PSI demonstrated significant positive effects on the reconstructed rigid-body kinematics of the knee, showing clearer coupled flexion/internal rotations - an important kinematic characteristic in healthy knees - than those without PSI during activities with or without weight-bearing. However, none of them showed normal contact patterns. The current findings will be helpful for surgical instrument design, as well as for surgical decision-making in MIS total knee arthroplasty.

Predicting Knee Joint Instability Using a Tibio-Femoral Statistical Shape Model

Statistical shape models (SSMs) are a well established computational technique to represent the morphological variability spread in a set of matching surfaces by means of compact descriptive quantities, traditionally called "modes of variation" (MoVs). SSMs of bony surfaces have been proposed in biomechanics and orthopedic clinics to investigate the relation between bone shape and joint biomechanics. In this work, an SSM of the tibio-femoral joint has been developed to elucidate the relation between MoVs and bone angular deformities causing knee instability. The SSM was built using 99 bony shapes (distal femur and proximal tibia surfaces obtained from segmented CT scans) of osteoarthritic patients. Hip-knee-ankle (HKA) angle, femoral varus-valgus (FVV) angle, internal-external femoral rotation (IER), tibial varus-valgus (TVV) angles, and tibial slope (TS) were available across the patient set. Discriminant analysis (DA) and logistic regression (LR) classifiers were adopted to underline specific MoVs accounting for knee instability. First, it was found that thirty-four MoVs were enough to describe 95% of the shape variability in the dataset. The most relevant MoVs were the one encoding the height of the femoral and tibial shafts (MoV #2) and the one representing variations of the axial section of the femoral shaft and its bending in the frontal plane (MoV #5). Second, using quadratic DA, the sensitivity results of the classification were very accurate, being all >0.85 (HKA: 0.96, FVV: 0.99, IER: 0.88, TVV: 1, TS: 0.87). The results of the LR classifier were mostly in agreement with DA, confirming statistical significance for MoV #2 (p = 0.02) in correspondence to IER and MoV #5 in correspondence to HKA (p = 0.0001), FVV (p = 0.001), and TS (p = 0.02). We can argue that the SSM successfully identified specific MoVs encoding ranges of alignment variability between distal femur and proximal tibia. This discloses the opportunity to use the SSM to predict potential misalignment in the knee for a new patient by processing the bone shapes, removing the need for measuring clinical landmarks as the rotation centers and mechanical axes.

Pair-wise vs group-wise registration in statistical shape model construction: representation of physiological and pathological variability of bony surface morphology

Statistical shape models (SSM) of bony surfaces have been widely proposed in orthopedics, especially for anatomical bone modeling, joint kinematic analysis, staging of morphological abnormality, and pre- and intra-operative shape reconstruction. In the SSM computation, reference shape selection, shape registration and point correspondence computation are fundamental aspects determining the quality (generality, specificity and compactness) of the SSM. Such procedures can be made critical by the presence of large morphological dissimilarities within the surfaces, not only because of anthropometrical variability but also mainly due to pathological abnormalities. In this work, we proposed a SW pipeline for SSM construction based on pair-wise (PW) shape registration, which requires the a-priori selection of the reference shape, and on a custom iterative point correspondence algorithm. We addressed large morphological deformations in five different bony surface sets, namely proximal femur, distal femur, patella, proximal fibula and proximal tibia, extracted from a retrospective patient dataset. The technique was compared to a method from the literature, based on group-wise (GW) shape registration. As a main finding, the proposed technique provided generalization and specificity median errors, for all the five bony regions, lower than 2 mm. The comparative analysis provided basically similar results. Particularly, for the distal femur that was the shape affected by the largest pathological deformations, the differences in generalization, specificity and compactness were lower than 0.5 mm, 0.5 mm, and 1%, respectively. We can argue the proposed pipeline, along with the robust correspondence algorithm, is able to compute high-quality SSM of bony shapes, even affected by large morphological variability.

A Novel Ultrasound-Based Lower Extremity Motion Tracking System

Tracking joint motion of the lower extremity is important for human motion analysis. In this study, we present a novel ultrasound-based motion tracking system for measuring three-dimensional (3D) position and orientation of the femur and tibia in 3D space and quantifying tibiofemoral kinematics under dynamic conditions. As ultrasound is capable of detecting underlying bone surface noninvasively through multiple layers of soft tissues, an integration of multiple A-mode ultrasound transducers with a conventional motion tracking system provides a new approach to track the motion of bone segments during dynamic conditions. To demonstrate the technical and clinical feasibilities of this concept, an in vivo experiment was conducted. For this purpose the kinematics of healthy individuals were determined in treadmill walking conditions and stair descending tasks. The results clearly demonstrated the potential of tracking skeletal motion of the lower extremity and measuring six-degrees-of-freedom (6-DOF) tibiofemoral kinematics and related kinematic alterations caused by a variety of gait parameters. It was concluded that this prototyping system has great potential to measure human kinematics in an ambulant, non-radiative, and noninvasive manner.

Validation of 2 noninvasive, markerless reconstruction techniques in biplane high-speed fluoroscopy for 3-dimensional research of bovine distal limb kinematics

Lameness severely impairs cattle's locomotion, and it is among the most important threats to animal welfare, performance, and productivity in the modern dairy industry. However, insight into the pathological alterations of claw biomechanics leading to lameness and an understanding of the biomechanics behind development of claw lesions causing lameness are limited. Biplane high-speed fluoroscopic kinematography is a new approach for the analysis of skeletal motion. Biplane high-speed videos in combination with bone scans can be used for 3-dimensional (3D) animations of bones moving in 3D space. The gold standard, marker-based animation, requires implantation of radio-opaque markers into bones, which impairs the practicability for lameness research in live animals. Therefore, the purpose of this study was to evaluate the comparative accuracy of 2 noninvasive, markerless animation techniques (semi-automatic and manual) in 3D animation of the bovine distal limb. Tantalum markers were implanted into each of the distal, middle, and proximal phalanges of 5 isolated bovine distal forelimbs, and biplane high-speed x-ray videos of each limb were recorded to capture the simulation of one step. The limbs were scanned by computed tomography to create bone models of the 6 digital bones, and 3D animation of the bones' movements were subsequently reconstructed using the marker-based, the semi-automatic, and the manual animation techniques. Manual animation translational bias and precision varied from 0.63 ± 0.26 mm to 0.80 ± 0.49 mm, and rotational bias and precision ranged from 2.41 ± 1.43° to 6.75 ± 4.67°. Semi-automatic translational values for bias and precision ranged from 1.26 ± 1.28 mm to 2.75 ± 2.17 mm, and rotational values varied from 3.81 ± 2.78° to 11.7 ± 8.11°. In our study, we demonstrated the successful application of biplane high-speed fluoroscopic kinematography to gait analysis of bovine distal limb. Using the manual animation technique, kinematics can be measured with sub-millimeter accuracy without the need for invasive marker implantation.

Statistical shape modeling predicts patellar bone geometry to enable stereo-radiographic kinematic tracking

Complications in the patellofemoral (PF) joint of patients with total knee replacements include patellar subluxation and dislocation, and remain a cause for revision. Kinematic measurements to assess these complications and evaluate implant designs require the accuracy of dynamic stereo-radiographic systems with 3D-2D registration techniques. While tibiofemoral kinematics are typically derived by tracking metallic implants, PF kinematic measurements are difficult as the patellar implant is radiotransparent and a representation of the resected patella bone requires either pre-surgical imaging and precise implant placement or post-surgical imaging. Statistical shape models (SSMs), used to characterize anatomic variation, provide an alternative means to obtain the representation of the resected patella for use in kinematic tracking. Using a virtual platform of a stereo-radiographic system, the objectives of this study were to evaluate the ability of an SSM to predict subject-specific 3D implanted patellar geometries from simulated 2D image profiles, and to formulate an effective data collection methodology for PF kinematics by considering accuracy for a variety of patient pose scenarios. An SSM of the patella was developed for 50 subjects and a leave-one-out approach compared SSM-predicted and actual geometries; average 3D errors were 0.45±0.07mm (mean±standard deviation), which is comparable to the accuracy of traditional segmentation. Further, initial imaging of the patella in five unique stereo radiographic perspectives yielded the most accurate representation. The ability to predict the remaining patellar geometry of the implanted PF joint with radiographic images and SSM, instead of CT, can reduce radiation exposure and streamline in vivo kinematic evaluations.

Three dimensional measurement of minimum joint space width in the knee from stereo radiographs using statistical shape models

Objectives

An important measure for the diagnosis and monitoring of knee osteoarthritis is the minimum joint space width (mJSW). This requires accurate alignment of the x-ray beam with the tibial plateau, which may not be accomplished in practice. We investigate the feasibility of a new mJSW measurement method from stereo radiographs using 3D statistical shape models (SSM) and evaluate its sensitivity to changes in the mJSW and its robustness to variations in patient positioning and bone geometry.

Materials and Methods

A validation study was performed using five cadaver specimens. The actual mJSW was varied and images were acquired with variation in the cadaver positioning. For comparison purposes, the mJSW was also assessed from plain radiographs. To study the influence of SSM model accuracy, the 3D mJSW measurement was repeated with models from the actual bones, obtained from CT scans.

Results

The SSM-based measurement method was more robust (consistent output for a wide range of input data/consistent output under varying measurement circumstances) than the conventional 2D method, showing that the 3D reconstruction indeed reduces the influence of patient positioning. However, the SSM-based method showed comparable sensitivity to changes in the mJSW with respect to the conventional method. The CT-based measurement was more accurate than the SSM-based measurement (smallest detectable differences 0.55 mm versus 0. 82 mm, respectively).

Conclusion

The proposed measurement method is not a substitute for the conventional 2D measurement due to limitations in the SSM model accuracy. However, further improvement of the model accuracy and optimisation technique can be obtained. Combined with the promising options for applications using quantitative information on bone morphology, SSM based 3D reconstructions of natural knees are attractive for further development.

Cite this article: E. A. van IJsseldijk, E. R. Valstar, B. C. Stoel, R. G. H. H. Nelissen, N. Baka, R. van’t Klooster, B. L. Kaptein. Three dimensional measurement of minimum joint space width in the knee from stereo radiographs using statistical shape models. Bone Joint Res 2016;320–327. DOI: 10.1302/2046-3758.58.2000626.

Constrained Statistical Modelling of Knee Flexion From Multi-Pose Magnetic Resonance Imaging

Reconstruction of the anterior cruciate ligament (ACL) through arthroscopy is one of the most common procedures in orthopaedics. It requires accurate alignment and drilling of the tibial and femoral tunnels through which the ligament graft is attached. Although commercial computer-assisted navigation systems exist to guide the placement of these tunnels, most of them are limited to a fixed pose without due consideration of dynamic factors involved in different knee flexion angles. This paper presents a new model for intraoperative guidance of arthroscopic ACL reconstruction with reduced error particularly in the ligament attachment area. The method uses 3D preoperative data at different flexion angles to build a subject-specific statistical model of knee pose. To circumvent the problem of limited training samples and ensure physically meaningful pose instantiation, homogeneous transformations between different poses and local-deformation finite element modelling are used to enlarge the training set. Subsequently, an anatomical geodesic flexion analysis is performed to extract the subject-specific flexion characteristics. The advantages of the method were also tested by detailed comparison to standard Principal Component Analysis (PCA), nonlinear PCA without training set enlargement, and other state-of-the-art articulated joint modelling methods. The method yielded sub-millimetre accuracy, demonstrating its potential clinical value.

Sequential 3D analysis of patellofemoral kinematics from biplanar x-rays: In vitro validation protocol

BACKGROUND

Developing criteria for assessing patellofemoral kinematics is crucial to understand, evaluate, and monitor patellofemoral function. The objective of this study was to assess a sequential 3D analysis method based on biplanar radiographs, using an in vitro protocol.

HYPOTHESIS

Biplanar radiography combined with novel 3D reconstruction methods provides a reliable evaluation of patellofemoral function, without previous imaging.

MATERIAL AND METHODS

Eight cadaver specimens were studied during knee flexion cycles from 0° to 60° induced by an in vitro simulator. The protocol was validated by investigating sequential and continuous motion using an optoelectronic system, evaluating measurement accuracy and reproducibility using metallic beads embedded in the patella, and comparing the 3D patellar geometry to computed tomography (CT) images.

RESULTS

The differences in position between the sequential and continuous kinematic analyses were less than 1mm and 1°. The protocol proved reliable for tracking several components of knee movements, including patellar translations, flexion, and tilt. In this analysis, uncertainty was less than 2 mm for translations and less than 3° for rotations, except rotation in the coronal plane. For patellar tilt, uncertainty was 5°. Mean difference in geometry was 0.49 mm.

DISCUSSION

Sequential analysis results are consistent with continuous kinematics. This analysis method provides patellar position parameters without requiring previous CT or magnetic resonance imaging. A clinical study may deserve consideration to identify patellofemoral kinematic profiles and position criteria in vivo.

LEVEL OF EVIDENCE

IV, experimental study.

Fluoroscopy-based tracking of femoral kinematics with statistical shape models

PURPOSE

Precise knee kinematics assessment helps to diagnose knee pathologies and to improve the design of customized prosthetic components. The first step in identifying knee kinematics is to assess the femoral motion in the anatomical frame. However, no work has been done on pathological femurs, whose shape can be highly different from healthy ones.

METHODS

We propose a new femoral tracking technique based on statistical shape models and two calibrated fluoroscopic images, taken at different flexion-extension angles. The cost function optimization is based on genetic algorithms, to avoid local minima. The proposed approach was evaluated on 3 sets of digitally reconstructed radiographic images of osteoarthritic patients.

RESULTS

It is found that using the estimated shape, rather than that calculated from CT, significantly reduces the pose accuracy, but still has reasonably good results (angle errors around 2[Formula: see text], translation around 1.5 mm).

Prediction of in vivo knee joint kinematics using a combined dual fluoroscopy imaging and statistical shape modeling technique

Using computed tomography (CT) or magnetic resonance (MR) images to construct 3D knee models has been widely used in biomedical engineering research. Statistical shape modeling (SSM) method is an alternative way to provide a fast, cost-efficient, and subject-specific knee modeling technique. This study was aimed to evaluate the feasibility of using a combined dual-fluoroscopic imaging system (DFIS) and SSM method to investigate in vivo knee kinematics. Three subjects were studied during a treadmill walking. The data were compared with the kinematics obtained using a CT-based modeling technique. Geometric root-mean-square (RMS) errors between the knee models constructed using the SSM and CT-based modeling techniques were 1.16 mm and 1.40 mm for the femur and tibia, respectively. For the kinematics of the knee during the treadmill gait, the SSM model can predict the knee kinematics with RMS errors within 3.3 deg for rotation and within 2.4 mm for translation throughout the stance phase of the gait cycle compared with those obtained using the CT-based knee models. The data indicated that the combined DFIS and SSM technique could be used for quick evaluation of knee joint kinematics.

Direct assessment of 3D foot bone kinematics using biplanar X-ray fluoroscopy and an automatic model registration method

BACKGROUND

Quantifying detailed 3-dimensional (3D) kinematics of the foot in contact with the ground during locomotion is crucial for understanding the biomechanical functions of the complex musculoskeletal structure of the foot. Biplanar X-ray fluoroscopic systems and model-based registration techniques have recently been employed to capture and visualise 3D foot bone movements in vivo, but such techniques have generally been performed manually. In the present study, we developed an automatic model-registration method with biplanar fluoroscopy for accurate measurement of 3D movements of the skeletal foot.

METHODS

Three-dimensional surface models of foot bones were generated prior to motion measurement based on computed tomography. The bone models generated were then registered to biplanar fluoroscopic images in a frame-by-frame manner using an optimisation technique, to maximise similarity measures between occluding contours of the bone surface models with edge-enhanced fluoroscopic images, while avoiding mutual penetration of bones. A template-matching method was also introduced to estimate the amount of bone translation and rotation prior to automatic registration.

RESULTS

We analysed 3D skeletal movements of a cadaver foot mobilized by a robotic gait simulator. The 3D kinematics of the calcaneus, talus, navicular and cuboid in the stance phase of the gait were successfully reconstructed and quantified using the proposed model-registration method. The accuracy of bone registration was evaluated as 0.27 ± 0.19 mm and 0.24 ± 0.19° (mean ± standard deviation) in translation and rotation, respectively, under static conditions, and 0.36 ± 0.19 mm and 0.42 ± 0.30° in translation and rotation, respectively, under dynamic conditions.

CONCLUSIONS

The measurement was confirmed to be sufficiently accurate for actual analysis of foot kinematics. The proposed method may serve as an effective tool for understanding the biomechanical function of the human foot during locomotion.

Segmentation of multiple heart cavities in 3-D transesophageal ultrasound images

Three-dimensional transesophageal echocardiography (TEE) is an excellent modality for real-time visualization of the heart and monitoring of interventions. To improve the usability of 3-D TEE for intervention monitoring and catheter guidance, automated segmentation is desired. However, 3-D TEE segmentation is still a challenging task due to the complex anatomy with multiple cavities, the limited TEE field of view, and typical ultrasound artifacts. We propose to segment all cavities within the TEE view with a multi-cavity active shape model (ASM) in conjunction with a tissue/blood classification based on a gamma mixture model (GMM). 3-D TEE image data of twenty patients were acquired with a Philips X7-2t matrix TEE probe. Tissue probability maps were estimated by a two-class (blood/tissue) GMM. A statistical shape model containing the left ventricle, right ventricle, left atrium, right atrium, and aorta was derived from computed tomography angiography (CTA) segmentations by principal component analysis. ASMs of the whole heart and individual cavities were generated and consecutively fitted to tissue probability maps. First, an average whole-heart model was aligned with the 3-D TEE based on three manually indicated anatomical landmarks. Second, pose and shape of the whole-heart ASM were fitted by a weighted update scheme excluding parts outside of the image sector. Third, pose and shape of ASM for individual heart cavities were initialized by the previous whole heart ASM and updated in a regularized manner to fit the tissue probability maps. The ASM segmentations were validated against manual outlines by two observers and CTA derived segmentations. Dice coefficients and point-to-surface distances were used to determine segmentation accuracy. ASM segmentations were successful in 19 of 20 cases. The median Dice coefficient for all successful segmentations versus the average observer ranged from 90% to 71% compared with an inter-observer range of 95% to 84%. The agreement against the CTA segmentations was slightly lower with a median Dice coefficient between 85% and 57%. In this work, we successfully showed the accuracy and robustness of the proposed multi-cavity segmentation scheme. This is a promising development for intraoperative procedure guidance, e.g., in cardiac electrophysiology.

Fully Automated Determination of Femoral Coordinate System in CT Image Based on Epicondyles

We propose a fully automated determination of the femoral coordinates in computerized tomography (CT) imaging based on epicondyles. The challenge point of this paper is that we take up how to calculate the femoral coordinate system (FCS), which is difficult to determine automatically. Our proposed method automatically determines the FCS based on anatomical reference points. We evaluated 10 subjects (six men and four women 28.9 ± 9.3 years old, three left-handed and seven right-handed) who had no history of joint injury. We examined the proposed method by comparing the expert and algorithm. The medial epicondyle was 1.41 ± 0.75 mmp= 0.42 > 0.05, student’sttest) in positioning accuracy. The lateral epicondyle was 1.36 ± 0.70 mmp= 0.42) in positioning accuracy. The origin was 0.87 ± 0.40 mmp= 0.71). in positioning accuracy. The lateral axis angle accuracy was 0.53 ± 0.84°p= 0.44). In short, the proposed method constructed patient-specific coordinate systems more accurately than expert manual.

Numerical optimization of alignment reproducibility for customizable surgical guides

Purpose

Computer-assisted orthopedic surgery aims at minimizing invasiveness, postoperative pain, and morbidity with computer-assisted preoperative planning and intra-operative guidance techniques, of which camera-based navigation and patient-specific templates (PST) are the most common. PSTs are one-time templates that guide the surgeon initially in cutting slits or drilling holes. This method can be extended to reusable and customizable surgical guides (CSG), which can be adapted to the patients’ bone. Determining the right set of CSG input parameters by hand is a challenging task, given the vast amount of input parameter combinations and the complex physical interaction between the PST/CSG and the bone.

Methods

This paper introduces a novel algorithm to solve the problem of choosing the right set of input parameters. Our approach predicts how well a CSG instance is able to reproduce the planned alignment based on a physical simulation and uses a genetic optimization algorithm to determine optimal configurations. We validate our technique with a prototype of a pin-based CSG and nine rapid prototyped distal femora.

Results

The proposed optimization technique has been compared to manual optimization by experts, as well as participants with domain experience. Using the optimization technique, the alignment errors remained within practical boundaries of 1.2 mm translation and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$0.9^\circ $$\end{document}0.9∘ rotation error. In all cases, the proposed method outperformed manual optimization.

Conclusions

Manually optimizing CSG parameters turns out to be a counterintuitive task. Even after training, subjects with and without anatomical background fail in choosing appropriate CSG configurations. Our optimization algorithm ensures that the CSG is configured correctly, and we could demonstrate that the intended alignment of the CSG is accurately reproduced on all tested bone geometries.