3D Analysis of Upper Limbs Motion during Rehabilitation Exercises Using the KinectTM Sensor: Development, Laboratory Validation and Clinical Application.

Optoelectronic devices are the gold standard for 3D evaluation in clinics, but due to the complexity of this kind of hardware and the lack of access for patients, affordable, transportable, and easy-to-use systems must be developed to be largely used in daily clinics. The KinectTM sensor has various advantages compared to optoelectronic devices, such as its price and transportability. However, it also has some limitations: (in)accuracy of the skeleton detection and tracking as well as the limited amount of available points, which makes 3D evaluation impossible. To overcome these limitations, a novel method has been developed to perform 3D evaluation of the upper limbs. This system is coupled to rehabilitation exercises, allowing functional evaluation while performing physical rehabilitation. To validate this new approach, a two-step method was used. The first step was a laboratory validation where the results obtained with the KinectTM were compared with the results obtained with an optoelectronic device; 40 healthy young adults participated in this first part. The second step was to determine the clinical relevance of this kind of measurement. Results of the healthy subjects were compared with a group of 22 elderly adults and a group of 10 chronic stroke patients to determine if different patterns could be observed. The new methodology and the different steps of the validations are presented in this paper.

Further consideration of the curvature of the Neandertal Femur.

OBJECTIVES: Neandertal femora are particularly known for having a marked sagittal femoral curvature. This study examined femoral curvature in Neandertals in comparison to a modern human population from Belgium by the use of three-dimensional (3D) quadric surfaces modeled from the bone surface. 3D models provide detailed information and enabled femoral curvature to be analyzed in conjunction with other morphological parameters. MATERIALS AND METHODS: 3D models were created from CT scans of 75 modern human femora and 7 Neandertal femora. Quadric surfaces (QS) were created from the triangulated surface vertices in all areas of interest (neck, head, diaphyseal shaft, condyles) extracted from previously placed anatomical landmarks. The diaphyseal shaft was divided into five QS shapes and curvature was measured by degrees of difference between QS shapes. Each bone was placed in a local coordinate system enabling each bone to be analyzed in the same way. RESULTS: The use of 3D quadric surface fitting allowed the distribution of curvature with similarly curved femora to be analyzed and the different patterns of curvature between the two groups to be determined. The Neandertals were shown to have a higher degree of femoral curvature and a more distal point of femoral curvature than the modern human population from Belgium. CONCLUSIONS: Morphological aspects of the Neandertal femur are different from this modern human population although mainly seem unrelated to femoral curvature. The relative lack of correlations with other femoral bony morphological factors suggests femoral curvature variations may be related to other aspects.

In-vivo analysis of sternal angle, sternal and sternocostal kinematics in supine humans during breathing.

This paper aims at contributing to the understanding of the combination of in vivo sternum displacement, sternal angle variations and sternocostal joints (SCJ) kinematics of the seven first rib pairs over the inspiratory capacity (IC). Retrospective codified spiral-CT data obtained at total lung capacity (TLC), middle of inspiratory capacity (MIC) and at functional residual capacity (FRC) were used to compute kinematic parameters of the bones and joints of interest in a sample of 12 asymptomatic subjects. 3D models of rib, thoracic vertebra, manubrium and sternum were processed to determine anatomical landmarks (ALs) on each bone. These ALs were used to create local coordinate system and compute spatial transformation of ribs and manubrium relative to sternum, and sternum relative to thoracic vertebra. The rib angular displacements and associated orientation of rotation axes and joint pivot points (JPP), the sternal angle variations and the associated displacement of the sternum relative to vertebra were computed between each breathing pose at the three lung volumes. Results can be summarized as following: (1) sternum cephalic displacement ranged between 17.8 and 19.2mm over the IC; (2) the sternal angle showed a mean variation of 4.4°±2.7° over the IC; (3) ranges of rib rotation relative to sternum decreased gradually with increasing rib level; (4) axes of rotation were similarly oriented at each SCJ; (5) JPP spatial displacements showed less variations at first SCJ compared to levels underneath; (6) linear relation was demonstrated between SCJ ROMs and sternum cephalic displacement over the IC.

How different are the Kebara 2 ribs to modern humans?

This study analyses rib geometric parameters of individual ribs of 14 modern human subjects (7 males and 7 females) in comparison to the reconstructed ribs of the Kebara 2 skeleton which was taken from the reconstruction of a Neandertal thorax by Sawyer & Maley (2005). Three-dimensional (3D) models were segmented from CT scans and each rib vertex cloud was placed into a local coordinate system defined from the rib principal axes. Rib clouds were then analysed using best fitting ellipses of the external contours of the cross-section areas. The centroid of each ellipse was then used to measure the centroidal pathway between each slice (rib midline). Curvature of the ribs was measured from the mid-line of the ribs as the sum of angles between successive centroids in adjacent cross sections. Distinct common patterns were noted in all rib geometric parameters for modern humans. The Kebara 2 reconstructed ribs also followed the same patterns. This study demonstrated that there are differences between the sexes in rib geometrical parameters, with females showing smaller rib width, chord length and arc length, but greater curvature (rib torsion, rib axial curvature, rib anterior-posterior bending) than males. The Kebara 2 reconstructed ribs were within the modern human range for the majority of geometrical parameters.

Relationship between costovertebral joint kinematics and lung volume in supine humans.

This study investigates the relationship between the motion of the first ten costovertebral joints (CVJ) and lung volume over the inspiratory capacity (IC) using detailed kinematic analysis in a sample of 12 asymptomatic subjects. Retrospective codified spiral-CT data obtained at total lung capacity (TLC), middle of inspiratory capacity (MIC) and at functional residual capacity (FRC) were analysed. CVJ 3D kinematics were processed using previously-published methods. We tested the influence of the side, CVJ level and lung volume on CVJ kinematics. In addition, the correlations between anthropologic/pulmonary variables and CVJ kinematics were analysed. No linear correlation was found between lung volumes and CVJ kinematics. Major findings concerning 3D kinematics can be summarized as follows: 1) Ranges-of-motion decrease gradually with increasing CVJ level; 2) rib displacements are significantly reduced at lung volumes above the MIC and do not differ between CVJ levels; 3) the axes of rotation of the ribs are similarly oriented for all CVJ levels.

Validation of the Balance Board for Clinical Evaluation of Balance During Serious Gaming Rehabilitation Exercises.

BACKGROUND: Balance and posture can be affected in various conditions or become decreased with aging. A diminution of balance control induces an increase of fall's risk. INTRODUCTION: The Nintendo Wii Balance Board™ (WBB) is used in rehabilitation to perform balance exercises (using commercial video games). The WBB has also been validated to assess balance and posture in static conditions. However, there is currently no study investigating the use of WBB to assess balance during the realization of balance exercises using this device. The aim of this study was to validate the use of WBB, coupled with specially developed serious games, to assess dynamic balance during rehabilitation exercises. MATERIALS AND METHODS: Thirty five subjects participated in this study. Subjects were asked to play two specially developed serious games. Center of pressure (CP) displacements were simultaneously recorded with a WBB and a gold standard force plate (FP). Nine parameters were derived from CP displacement. Bland and Altman plots, paired-sample t tests, intraclass correlation coefficient's, and Pearson's coefficient correlations were computed. RESULTS: Excellent correlation between both devices was found for each parameter for the two games (R = 0.95 and 0.96). DISCUSSION: Unlike previous work on the WBB, these excellent results were obtained without using any calibration procedure. Despite this, results were highly correlated between the WBB and the FP. CONCLUSIONS: The WBB could be used in clinics to assess balance during rehabilitation exercises and, thus, allows a more regular patient follow-up.

Femoral curvature variability in modern humans using three-dimensional quadric surface fitting.

INTRODUCTION: This study analysed femoral curvature in a population from Belgium in conjunction with other morphological characteristics by the use of three-dimensional (3D) quadric surfaces (QS) modelled from the bone surface. METHODS: 3D models were created from computed tomography data of 75 femoral modern human bones. Anatomical landmarks (ALs) were palpated in specific bony areas of the femur (shaft, condyles, neck and head). QS were then created from the surface vertices which enclose these ALs. The diaphyseal shaft was divided into five QS shapes to analyse curvature in different parts of the shaft. RESULTS: Femoral bending differs in different parts of the diaphyseal shaft. The greatest degree of curvature was found in the distal shaft (mean 4.5° range 0.2°-10°) followed by the proximal (mean 4.4° range 1.5°-10.2°), proximal intermediate (mean 3.7° range 0.9°-7.9°) and distal intermediate (mean 1.8° range 0.2°-5.6°) shaft sections. The proximal and distal angles were significantly more bowed than the intermediate proximal and the intermediate distal angle. There was no significant difference between the proximal and distal angle. No significant correlations were found between morphological characteristics and femoral curvature. An extremely large variability of femoral curvature with several bones displaying very high or low degrees of femoral curvature was also found. CONCLUSION: 3D QS fitting enables the creation of accurate models which can discriminate between different patterns in similar curvatures and demonstrates there is a clear difference between curvature in different parts of the shaft.

Effect of anatomical landmark perturbation on mean helical axis parameters of in vivo upper costovertebral joints.

The literature concerning quantification of costovertebral joint (CVJ) motion under in vivo conditions is scarce. Most papers concerning this topic are related to ex vivo loading conditions. In vivo protocols are available from the literature to determine rib and vertebra kinematics but new developments are needed to improve data processing concerning CVJ behaviour obtained from discrete breathing positions showing limiting ranges-of-motion and sensitive to noise. Data from previous work were used to implement a method analyzing mean helical axis (MHA) and pivot point parameters of the CVJ complexes. Several levels of noises were estimated within Monte-Carlo simulations to optimize MHA results. MHA parameters were then used to transform and define a CVJ-specific local coordinate system. This study proposes an improvement for CVJ kinematics processing and description from in vivo data obtained from computed tomography. This methodology emphasizes the possibility to work with variability of MHA parameters using Monte-Carlo procedures on anatomical landmark coordinates and to define a local coordinate system from this particular joint behaviour. Results from the CVJ joint model are closer to a hinge joint (secondary motions inferior to 3°) when anatomical frames are expressed from MHA orientation. MHA orientation and position data obtained from the proposed method are relevant according to angular dispersion obtained (from 7.5° to 13.9°) and therefore relevant to define behaviour of CVJ.

Physiologically corrected coupled motion during gait analysis using a model-based approach.

Gait analysis is used in daily clinics for patients' evaluation and follow-up. Stereophotogrammetric devices are the most used tool to perform these analyses. Although these devices are accurate results must be analyzed carefully due to relatively poor reproducibility. One of the major issues is related to skin displacement artifacts. Motion representation is recognized reliable for the main plane of motion displacement, but secondary motions, or combined, are less reliable because of the above artifacts. Model-based approach (MBA) combining accurate joint kinematics and motion data was previously developed based on a double-step registration method. This study presents an extensive validation of this MBA method by comparing results with a conventional motion representation model. Thirty five healthy subjects participated to this study. Gait motion data were obtained from a stereophotogrammetric system. Plug-in Gait model (PiG) and MBA were applied to raw data, results were then compared. Range-of-motion, were computed for pelvis, hip, knee and ankle joints. Differences between PiG and MBA were then computed. Paired-sample t-tests were used to compare both methods. Normalized root-mean square errors were also computed. Shapes of the curves were compared using coefficient of multiple correlations. The MBA and PiG approaches shows similar results for the main plane of motion displacement but statistically significative discrepancies appear for the combined motions. MBA appear to be usable in applications (such as musculoskeletal modeling) requesting better approximations of the joints-of-interest thanks to the integration of validated joint mechanisms.

Determination of repeatability of kinect sensor.

BACKGROUND: The Kinect™ (Microsoft™, Redmond, WA) sensor, originally developed for gaming purposes, may have interesting possibilities for other fields such as posture and motion assessment. The ability of the Kinect sensor to perform biomechanical measurements has previously been studied and shows promising results. However, interday repeatability of the device is still not known. MATERIALS AND METHODS: This study assessed the intra- and interday repeatability of the Kinect sensor compared with a standard stereophotogrammetric device during posture assessment for measuring segment lengths. Forty subjects took part in the study. Five motionless captures were performed in one session to assess posture. Data were simultaneously recorded with both devices. RESULTS: Similar intraclass correlations coefficient (ICC) values were found for intraday (ICC=0.94 for the Kinect device and 0.98 for the stereophotogrammetric device) and interday (ICC=0.88 and 0.87, respectively) repeatability. CONCLUSIONS: Results of this study suggest that a cost-effective, easy-to-use, and portable single markerless camera offers the same repeatability during posture assessment as an expensive, time-consuming, and nontransportable marker-based device.

In vivo thorax 3D modelling from costovertebral joint complex kinematics.

BACKGROUND: The costovertebral joint complex is mechanically involved in both respiratory function and thoracic spine stability. The thorax has been studied for a long time to understand its involvement in the physiological mechanism leading to specific gas exchange. Few studies have focused on costovertebral joint complex kinematics, and most of them focused on experimental in vitro analysis related to loading tests or global thorax and/or lung volume change analysis. There is however a clinical need for new methods allowing to process in vivo clinical data. This paper presents results from in vivo analysis of the costovertebral joint complex kinematics from clinically-available retrospective data. METHODS: In this study, in vivo spiral computed tomography imaging data were obtained from 8 asymptomatic subjects at three different lung volumes (from total lung capacity to functional residual capacity) calibrated using a classical spirometer. Fusion methods including 3D modelling and kinematic analysis were used to provide 3D costovertebral joint complex visualization for the true ribs (i.e., first seven pairs of ribs). FINDINGS: The 3D models of the first seven pairs of costovertebral joint complexes were obtained. A continuous kinematics simulation was interpolated from the three discrete computerized tomography positions. Helical axis representation was also achieved. INTERPRETATION: Preliminary results show that the method leads to meaningful and relevant results for clinical and pedagogical applications. Research in progress compares data from a sample of healthy volunteers with data collected from patients with cystic fibrosis to obtain new insights about the costovertebral joint complex range of motion and helical axis assessment in different pathological conditions.

Sex determination using the Probabilistic Sex Diagnosis (DSP: Diagnose Sexuelle Probabiliste) tool in a virtual environment.

The hip bone is one of the most reliable indicators of sex in the human body due to the fact it is the most dimorphic bone. Probabilistic Sex Diagnosis (DSP: Diagnose Sexuelle Probabiliste) developed by Murail et al., in 2005, is a sex determination method based on a worldwide hip bone metrical database. Sex is determined by comparing specific measurements taken from each specimen using sliding callipers and computing the probability of specimens being female or male. In forensic science it is sometimes not possible to sex a body due to corpse decay or injury. Skeletalization and dissection of a body is a laborious process and desecrates the body. There were two aims to this study. The first aim was to examine the accuracy of the DSP method in comparison with a current visual sexing method on sex determination. A further aim was to see if it was possible to virtually utilise the DSP method on both the hip bone and the pelvic girdle in order to utilise this method for forensic sciences. For the first part of the study, forty-nine dry hip bones of unknown sex were obtained from the Body Donation Programme of the Université Libre de Bruxelles (ULB). A comparison was made between DSP analysis and visual sexing on dry bone by two researchers. CT scans of bones were then analysed to obtain three-dimensional (3D) virtual models and the method of DSP was analysed virtually by importing the models into a customised software programme called lhpFusionBox which was developed at ULB. The software enables DSP distances to be measured via virtually-palpated bony landmarks. There was found to be 100% agreement of sex between the manual and virtual DSP method. The second part of the study aimed to further validate the method by analysing thirty-nine supplementary pelvic girdles of known sex blind. There was found to be a 100% accuracy rate further demonstrating that the virtual DSP method is robust. Statistically significant differences were found in the identification of sex between researchers in the visual sexing method although both researchers identified the same sex in all cases in the manual and virtual DSP methods for both the hip bones and pelvic girdles.

[A technological platform for cerebral palsy - the ICT4Rehab project]. / Une plate-forme technologique liée à la paralysie cérébrale - le projet ICT4Rehab.

The musculoskeletal system (MSS) is essential to allow us performing every-day tasks, being able to have a professional life or developing social interactions with our entourage. MSS pathologies have a significant impact on our daily life. It is therefore not surprising to find MSS-related health problems at the top of global statistics on professional absenteeism or societal health costs. The MSS is also involved in central nervous conditions, such as cerebral palsy (CP). Such conditions show complex etiology that complicates the interpretation of the observable clinical signs and the establishment of a wide consensus on the best practices to adopt for clinical monitoring and patient follow-up. These elements justify the organization of fundamental and applied research projects aiming to develop new methods to help clinicians to cope with the complexity of some MSS disorders. The ICT4Rehab project (www.ict4rehab.org) developed an integrated platform providing tools that enable easier management and visualization of clinical information related to the MSS of CP patients. This platform is opened to every interested clinical centre.

Musculoskeletal modeling of the suboccipital spine: kinematics analysis, muscle lengths, and muscle moment arms during axial rotation and flexion extension.

STUDY DESIGN: In vitro and modeling study of upper cervical spine (UCS) three-dimensional (3D) kinematics and muscle moment arm (MA) during axial rotation (AR) and flexion extension (FE). OBJECTIVE: To create musculoskeletal models with movement simulation including helical axis (HA) and muscle features. SUMMARY OF BACKGROUND DATA: Integration of various kinematics and muscle data into specific-specimen 3D anatomical models with graphical representation of HA and muscle orientation and MA is not reported for the UCS musculoskeletal system. METHODS: Kinematics, anatomical, and computed tomographic imaging data were sampled in 10 anatomical specimens. Using technical markers and anatomical landmarks digitizing, spatial position of segments was computed for five discrete positions of AR and FE using a 3D digitizer. To obtain musculoskeletal model simulation, a registration method was used to combine collected data. Processing was performed using orientation vector and HA computation and suboccipital muscle features (i.e., length and MA) relative to motion angle. RESULTS: Range of motion and coupling were in agreement with previous in vitro studies. HA (i.e., location and orientation) showed low variation at the occipitoaxial and atlantoaxial levels for FE and AR, respectively. The main orientation of the HA was vertical at C1-C2 during AR and horizontal at C0-C1 during FE. For muscles MA, absolute peak value (ranging from 20 to 40 mm) occurred at different poses depending on the analyzed muscle and motion. Poor magnitude was found for obliquus capitis inferior and rectus capitis posterior minor in FE and AR, respectively. CONCLUSION: On the basis of previous methods, we developed a protocol to create UCS musculoskeletal modeling with motion simulation including HA and suboccipital muscles representation. In this study, simultaneous segmental movement displaying with HA and muscles features was shown to be feasible.

In vitro 3D-kinematics of the upper cervical spine: helical axis and simulation for axial rotation and flexion extension.

PURPOSE: Registration of 3D-anatomical model and kinematics data is reported to be an accurate method to provide 3D-joint simulation. We applied this approach to discrete kinematics analysis of upper cervical spine (UCS) during axial rotation (AR) and flexion extension (FE) to create anatomical models with movement simulation including helical axis. METHODS: Kinematics and CT imaging data were sampled in ten anatomical specimens. Using technical and anatomical marker digitizing, spatial position of segments was computed for five discrete positions of AR and FE using a 3D-digitizer. Computerized tomography was used to create anatomical models and to assure kinematics and imaging data registration for simulation. Kinematics was processed using orientation vector and helical axis (HA) computation. RESULTS: Maximal standard error on marker digitizing was 0.47 mm. Range of motion and coupled movement during AR was in agreement with previous in vitro studies. HA location and orientation have shown low variation at the occipitoaxial and atlantoaxial levels for FE and AR, respectively. CONCLUSIONS: We developed a protocol to create UCS anatomical model simulations including three-dimensional discrete kinematics, using previously validated methods. In this study, simultaneous segmental movement simulation and display of HA variations was shown to be feasible. Although partially confirming previous results, helical axis computation showed variations of motion patterns dependent on movement, level and specimen. Further in vivo investigations are needed to confirm relevance of this method in the clinical field.

In vivo registration of both electrogoniometry and medical imaging: development and application on the ankle joint complex.

An in vivo method for joint kinematics visualization and analysis is described. Low-dose computed tomography allowed three-dimensional joint modeling, and electrogoniometry collected joint kinematic data. Data registration occurred using palpated anatomical landmarks to obtain interactive computer joint simulation. The method was applied on one volunteer's ankle, and reproducibility was tested (maximal discrepancy: 3.6 deg and 5.5 mm for rotation and translation respectively).

Double-step registration of in vivo stereophotogrammetry with both in vitro 6-DOFs electrogoniometry and CT medical imaging.

Standard registration techniques of bone morphology to motion analysis data often lead to unsatisfactory motion simulation because of discrepancies during the location of anatomical landmarks in the datasets. This paper describes an iterative registration method of a three-dimensional (3D) skeletal model with both 6 degrees-of-freedom joint kinematics and standard motion analysis data. The method is demonstrated in this paper on the lower limb. The method includes two steps. A primary registration allowed synchronization of in vitro kinematics of the knee and ankle joints using flexion/extension angles from in vivo gait analysis. Results from primary registration were then improved by a so-called advanced registration, which integrated external constraints obtained from experimental gait pre-knowledge. One cadaver specimen was analyzed to obtain both joint kinematics of knee and ankle joints using 3D electrogoniometry, and 3D bone morphology from medical imaging data. These data were registered with motion analysis data from a volunteer during the execution of locomotor tasks. Computer graphics output was implemented to visualize the results for a motion of sitting on a chair. Final registration results allowed the observation of both in vivo motion data and joint kinematics from the synchronized specimen data. The method improved interpretation of gait analysis data, thanks to the combination of realistic 3D bone models and joint mechanism. This method should be of interest both for research in gait analysis and medical education. Validation of the overall method was performed using RMS of the differences between bone poses estimated after registration and original data from motion analysis.

Calibration and validation of 6 DOFs instrumented spatial linkage for biomechanical applications. A practical approach.

A method for both calibration and validation of a 6 DOF electrogoniometer is presented. A 6 Revolute Instrumented Spatial Linkage (6R-ISL) and a three-dimensional digitizer (3DD) were used simultaneously to collect both static and continuous poses of unconstrained or constrained motions. Validation occurred using a calibrated ball-and-socket joint. A parametrical model of the 6R-ISL (i.e. Virtual Goniometer or VG) was designed using a standard multibody system geometry. Two approaches were used to adjust the VG parameters: a parametrical adjustment of the VG linkage geometry, and a functional adjustment of the potentiometer calibration curves (angle-voltage) in a predefined range of motion. After calibration, 6R-ISL accuracy was better than 1 mm and 1 degrees for translation and orientation, respectively. The functional method presented in this paper can be suggested as a practical approach, which allows on-line checking and calibration of 6R-ISL within the specific range of interest of a particular anatomical joint. In addition, improving the potentiometer calibration curves was less time consuming than the parametrical adjustment.

Development of multimedia learning modules for teaching human anatomy: application to osteology and functional anatomy.

Computer-assisted learning (CAL) is growing quickly within academic programs. Although the anatomical commercial packages that are available for this learning have attractive advantages, they also have drawbacks: they are frequently not in the local language of the students, they do not perfectly answer the needs of the local academic program, and their cost is frequently more than students can afford. This study describes a relatively inexpensive method to create CAL tutorials, whose content can be fully customized to local academic needs in terms of both program and language. The study describes its use in creating multimedia learning modules (MLMs) about Osteology and joint kinematics. The pedagogical content in these modules was collected from objective experiments to give students the opportunity to access new scientific knowledge during their education. It can be replaced, as desired, by almost any content due to the flexibility of the production method. Each MLM consists of two complementary subelements: a multimedia theoretical lecture and a three-dimensional interactive laboratory. Such MLMs are in use at both the University of Brussels (ULB) and the National University of Rwanda (NUR). The development of this work was part of the VAKHUM project, and the pedagogical validation is currently being performed as part of the MULTIMOD project.

Data representation for joint kinematics simulation of the lower limb within an educational context.

Three-dimensional (3D) visualization is becoming increasingly frequent in both qualitative and quantitative biomechanical studies of anatomical structures involving multiple data sources (e.g. morphological data and kinematics data). For many years, this kind of experiment was limited to the use of bi-dimensional images due to a lack of accurate 3D data. However, recent progress in medical imaging and computer graphics has forged new perspectives. Indeed, new techniques allow the development of an interactive interface for the simulation of human motions combining data from both medical imaging (i.e., morphology) and biomechanical studies (i.e., kinematics). Fields of application include medical education, biomechanical research and clinical research. This paper presents an experimental protocol for the development of anatomically realistic joint simulation within a pedagogical context. Results are shown for the lower limb. Extension to other joints is straightforward. This work is part of the Virtual Animation of the Kinematics of the Human project (VAKHUM) (http://www.ulb.ac.be/project/vakhum).