Evaluation of suitability of a micro-processing unit of motion analysis for upper limb tracking

The aim of this study is to assess the suitability of a micro-processing unit of motion analysis (MPUMA), for monitoring, reproducing, and tracking upper limb movements. The MPUMA is based on an inertial measurement unit, a 16-bit digital signal controller and a customized algorithm. To validate the performance of the system, simultaneous recordings of the angular trajectory were performed with a video-based motion analysis system. A test of the flexo-extension of the shoulder joint during the active elevation in a complete range of 120º of the upper limb was carried out in 10 healthy volunteers. Additional tests were carried out to assess MPUMA performance during upper limb tracking. The first, a 3D motion reconstruction of three movements of the shoulder joint (flexo-extension, abduction–adduction, horizontal internal–external rotation), and the second, an upper limb tracking online during the execution of three movements of the shoulder joint followed by a continuous random movement without any restrictions by using a virtual model and a mechatronic device of the shoulder joint. Experimental results demonstrated that the MPUMA measured joint angles that are close to those from a motion-capture system with orientation RMS errors less than 3º.     

Use of multiple wearable inertial sensors in upper limb motion tracking

This paper presents a new human motion tracking system using two wearable inertial sensors that are placed near the wrist and elbow joints of the upper limb. Each inertial sensor consists of a tri-axial accelerometer, a tri-axial gyroscope and a tri-axial magnetometer. The turning rates of the gyroscope were utilised for localising the wrist and elbow joints on the assumption that the two upper limb segment lengths are known a priori. To determine the translation and rotation of the shoulder joint, an equality-constrained optimisation technique is adopted to find an optimal solution, incorporating measurements from the tri-axial accelerometer and gyroscope. Experimental results demonstrate that this new system, compared to an optical motion tracker, has RMS position errors that are normally less than 0.01 m, and RMS angle errors that are 2.5-4.8 degrees .     

Perception of forearm angles in 3-dimensional space

Summary The purpose of this study was to determine a preferred coordinate system for representation of forearm orientation in 3-dimensional space. In one experiment, the ability of human subjects to perceive angles of the forearm in 3-dimensional space (forearm elevation and yaw — extrinsic coordinate system) was compared to their ability to perceive elbow joint angle (intrinsic coordinate system). While blindfolded, subjects performed an angle reproduction task in which the experimenter first positioned the upper limb in a reference trial. This was followed, after movement of the subject's entire upper limb to a different position, by an attempt to reproduce or match a criterior angle of the reference trial by motion of the forearm in elbow flexion or extension only. Note that matching of the criterion forearm angle in the new upper limb position could not be accomplished by reproducing the entire reference upper limb position, but only by angular motion at the elbow. Matching of all 3 criterion angles was accomplished with about equal accuracy in terms of absolute constant errors and variable errors. Correlation analysis of the perceptual errors showed that forearm elevation and elbow angle perception errors were not biased but that forearm yaw angle matching showed a bias toward elbow angle matching in 7 of 9 subjects. That is, errors in forearm yaw perception were attributed to a tendency toward a preferred intrinsic coordinate system for perception of forearm orientation. These results show that subjects can accurately perceive angles in both extrinsic and intrinsic coordinate systems in 3-dimensional space. Thus, these data conflict with previous reports of highly inaccurate perception of elbow joint angles in comparison to perception of forearm elevation. In an attempt to resolve this conflict with previous results, a second experiment was carried out in which perception of forearm elevation and elbow joint angles with the forearm motion constrained to a vertical plane. Results of this experiment showed that during a two-limb elbow angle matching task, four of five subjects exhibited a clear bias toward forearm elevation angle. During a one-limb angle reproduction task only two of five subjects exhibited such a bias. Perception of elevation angles show little bias toward elbow angle matching. These results indicate that use of tasks in which the limb is supported against gravity and motion is constrained to a vertical plane cause subjects to make perceptual errors during elbow angle matching such that the slopes of the forearms in a vertical plane (elevation angles) are more easily matched. It is concluded that human subjects can use both extrinsic and intrinsic coordinate systems in planning movements. Kinematic aspects may be planned in terms of an extrinsic coordinate system because of the use of vision in specifying location of external targets, but kinetic aspects of movement planning probably requires use of both forearm elevation angles and elbow joint angles to accurately specify forces and torques for muscles spanning the elbow.     

Gaze direction effects on perceptions of upper limb kinesthetic coordinate system axes

The effects of varying gaze direction on perceptions of the upper limb kinesthetic coordinate system axes and of the median plane location were studied in nine subjects with no history of neuromuscular disorders. In two experiments, six subjects aligned the unseen forearm to the trunk-fixed anterior-posterior (a/p) axis and earth-fixed vertical while gazing at different visual targets using either head or eye motion to vary gaze direction in different conditions. Effects of support of the upper limb on perceptual errors were also tested in different conditions. Absolute constant errors and variable errors associated with forearm alignment to the trunk-fixed a/p axis and earth-fixed vertical were similar for different gaze directions whether the head or eyes were moved to control gaze direction. Such errors were decreased by support of the upper limb when aligning to the vertical but not when aligning to the a/p axis. Regression analysis showed that single trial errors in individual subjects were poorly correlated with gaze direction, but showed a dependence on shoulder angles for alignment to both axes. Thus, changes in position of the head and eyes do not influence perceptions of upper limb kinesthetic coordinate system axes. However, dependence of the errors on arm configuration suggests that such perceptions are generated from sensations of shoulder and elbow joint angle information. In a third experiment, perceptions of median plane location were tested by instructing four subjects to place the unseen right index fingertip directly in front of the sternum either by motion of the straight arm at the shoulder or by elbow flexion/extension with shoulder angle varied. Gaze angles were varied to the right and left by 0.5 radians to determine effects of gaze direction on such perceptions. These tasks were also carried out with subjects blind-folded and head orientation varied to test for effects of head orientation on perceptions of median plane location. Constant and variable errors for fingertip placement relative to the sternum were not affected by variations in gaze direction or head orientation. Thus, the perceived position of the trunk-fixed median plane is not altered by varying gaze direction. The implications of these results for mechanisms underlying kinesthetic perceptions and their potential roles in programming of upper limb movements to visual targets are discussed.     

Transformations between visual and kinesthetic coordinate systems in reaches to remembered object locations and orientations

The abilities of human subjects to perform reach and grasp movements to remembered locations/ orientations of a cylindrical object were studied under four conditions: (1) visual presentation of the object — reach with vision allowed; (2) visual presentation — reach while blindfolded; (3) kinesthetic presentation of the object-reach while blindfolded and (4) kinesthetic presentation-reach with vision. The results showed that subjects were very accurate in locating the object in the purely kinesthetic condition and that directional errors were low in all four conditions; but, predictable errors in reach distance occurred in conditions 1,2, and 4. The pattern of these distance errors was similar to that identified in previous research using a pointing task to a small target (i.e., overshoots of close targets, undershoots of far targets). The observation that the pattern of distance errors in condition 4 was similar to that of conditions 1 and 2 suggests that subjects transform kinesthetically defined hand locations into a visual coordinate system when vision is available during upper limb motion to a remembered kinesthetic target. The differences in orientation of the upper limb between target and reach positions in condition 3 were similar in magnitude to the errors associated with kinesthetic perceptions of arm and hand orientations in three-dimensional space reported in previous studies. However, fingertip location was specified with greater accuracy than the orientation of upper limb segments. This was apparently accomplished by compensation of variations in shoulder (arm) angles with oppositely directed variations in elbow joint angles. Subjects were also able to transform visually perceived object orientation into an appropriate hand orientation for grasp, as indicated by the relation between hand roll angle and object orientation (elevation angle). The implications of these results for control of upper limb motion to external targets are discussed.     

一种简单的上肢关节坐标系定义方法

人体运动的运动学或者动力学分析通常需要对各关节建立局部坐标系,目前大部分人体关节坐标系定义方法基于临床应用,要求专业的解剖学知识,可操作性差.本文定义了上肢的主要解剖学特征点,通过这些特征点构建了肩关节、肘关节及腕关节坐标系,并详细阐述了直接采用肩锁关节点（AC）构建肩关节坐标系的方法和必要性.利用运动捕捉系统对自愿者步态过程中的上肢运动轨迹进行采集,通过分析坐标系特征点在运动过程中空间相对位置的不变性,说明了文章提出的上肢局部坐标系定义方法在一定范围内是合理的.     

Limb segment inclination sense in proprioception

Summary Two experiments were performed to determine if proprioceptive signals are perceived more readily in terms of limb segment inclinations to the vertical than as joint angles. Subjects attempted to match arm positions with the upper arms supported at different inclinations. Constant error data showed that, when instructed to match forearm inclinations to the vertical, subjects were very accurate. When required to match elbow joint angles, however, errors were strongly biased in the direction of matching forearm inclinations. The results support a view of proprioception as a system in which afferent signals related to the gravitational torques acting at joints lead to the perception of limb orientation rather than joint angles. Such a system would allow more efficient determination of the relationship of limb segments to external objects than would one based purely on joint angles.This work was supported in part by research awards from the U.S. Public Health Service, AG05154, NS17421 and NATO no 227/82     

Effect of different inertial parameter sets on joint moment calculation during stair ascending and descending

The reliability of internal joint moment calculation in gait analysis during daily living activities is fundamental for clinical decisions based on joint function. This calculation, obtained by means of the inverse dynamics, depends on several modelling factors, such as assumptions on the segments and on the relevant joints constituting the kinematic chain. In this study, the effect of five different sets of inertial parameters on three-dimensional calculation of lower limb joint moments was investigated during the stair ascending and descending of 10 young subjects. The lower limb was represented as a chain of three rigid segments: foot, shank and thigh. The inertial parameters sets were taken from the literature. The root mean square value over the step cycle of the difference between joint moments calculated at the lower limb with different inertial parameter sets expressed in percentage of their corresponding range was computed. The results showed small differences between ex vivo and in vivo data, between data from different populations and among different modality of inertial parameters acquisition. The root mean square value was negligible at the ankle and increased as moving proximally among the joints: the maximum was 21.8% in the internal/external rotation moment at the hip. In order to achieve accurate estimate of lower limb joint moments other factors should be investigated rather than optimal inertial parameter set.     

抓举过程中运动员的上肢运动学分析

人体运动过程中,身体各部位的加速度、角加速度等运动学参数对计算关节力、关节力矩及肌肉力等动力学参数起着至关重要的作用.文章通过分析人体运动过程中关节坐标系相对世界坐标系的瞬时位置关系,以及各时刻关节坐标系空间位置关系,提出了基于关节坐标系的人体运动学参数计算方法.然后,以运动员抓举杠铃过程中的上肢运动为例,利用运动捕捉系统进行数据采集,计算出前臂在运动过程中的质心平移加速度和角加速度.结果 表明,前臂质心总的平移加速度主要由竖直方向的分量构成,而角加速度主要在改变杠铃轨迹过程中产生.     

基于多模态信息融合的肘关节连续运动估计

目的 针对目前上肢康复训练设备多为被动式、训练方式单一、患者主动参与度较低等问题,提出一种基于多模态信息融合的上肢连续运动估计算法,实现对肘关节力矩的准确估计。方法 首先,在4种角速度下,采集受试者的表面肌电信号和姿态信号,提取信号的时域特征并利用主成分分析方法进行特征融合;其次,通过附加动量法和自适应学习率对反向传播神经网络(back propagation neural network, BPNN)进行改进,使用粒子群算法(particle swarm optimization, PSO)对神经网络进行优化,构建基于PSO-BPNN的连续运动估计模型;最后,以第2类拉格朗日方程计算的关节力矩作为准确值,对模型进行训练,并与传统BPNN模型进行性能对比。结果 传统BP神经网络模型均方根误差为558.9 mN·m,R²系数为77.19%,优化模型后的均方根误差和R²系数分别为113.6 N·m、99.12%,力矩估计准确度进一步提高。结论 本文提出的肘关节连续运动估计方法能够准确地识别运动意图,为上肢外骨骼康复机器人的主动控制提供切实可行的方...     

Development of a powered variable-stiffness exoskeleton device for elbow rehabilitation

Robot-assisted movement training by means of exoskeleton devices has been proven to be an effective method for post-stroke patients to recover their motor function. However, in order to be used in home-based rehabilitation, the kinematic structure of a wearable exoskeleton device should provide portability and make allowances for the natural joint range of motion for the user. Additionally, the actuated stiffness of the target joint is desired to be adjustable in accordance with the specific impairment level of the patient’s upper limb. In this paper, we present a novel portable exoskeleton device which could provide support for rehabilitation patients with variable actuated stiffness in the elbow joint. It has five passive degrees of freedom to guarantee the user’s natural joint range of motion and intra-subject variability, as well as an integrated variable stiffness actuator (VSA) which can adjust the joint stiffness independently by moving the pivot position. An elbow power-assist trial with different actuated joint stiffnesses was tested on a healthy subject to evaluate the functionality of the proposed device. By regulating the joint stiffness, the proposed device could provide variable power assistance for the wearer’s elbow movements.     

Apparatus to measure simultaneously 14 isometric leg joint moments. Part 2: Multi-moment chair system

An apparatus has been developed that measures isometrically the 14 lower limb joint moments corresponding to the degrees of freedom of the hips, knees and ankles. This is the second of two papers describing the development of the multi-moment chair system (MMCS). It presents the overall design and changes that were implemented to compensate for problems. These were primarily to improve the accuracy of hip joint moments; a compromise between accuracy and practicalities, because of force-moment responses being measured at the ankles. All joint moment errors have been calculated to be of the order of a few newton metres. Since these represent errors of less than 10% when cosindering peak joint moment responses, this is considered sufficiently accurate for the proposed application. The MMCS is being used in a programme to investigate the restoration of lower limb functions, principally standing, in paraplegics by electrical stimulation of the lumbosacral anterior roots.     

An optimization-based simultaneous approach to the determination of muscular, ligamentous, and joint contact forces provides insight into musculoligamentous interaction

Typical inverse dynamics approaches to the calculation of muscle, ligament, and joint contact forces are based on a step-wise solution of the equations of motion. This approach is therefore limited in its ability to provide insight as to the muscular, ligamentous, and articular interactions that create joint stability. In this study, a new musculoskeletal model of the lower limb is described, in which the equations of motion describing the force and moment equilibrium at the joints of the lower limb are solved simultaneously using optimization techniques. The new model was employed to analyze vertical jumping using a variety of different optimization cost functions and the results were compared to more traditional approaches. The new model was able to find a solution with lower muscular force upper bounds due to the ability of the ligaments to contribute to moment equilibrium at the ankle and knee joints. Equally, the new model produced lower joint contact forces than traditional approaches for cases which also included a consideration as to ligament or joint contact forces within the cost function. This study demonstrates the possibility of solving the inverse dynamic equations of motion simultaneously using contemporary technology, and further suggests that this might be important due to the complementary function of the muscles and ligaments in creating joint stability.     

Apparatus to measure simultaneously 14 isometric leg joint moments. Part 2: Multi-moment chair system

An apparatus has been developed that measures isometrically the 14 lower limb joint moments corresponding to the degrees of freedom of the hips, knees and ankles. This is the second of two papers describing the development of the multi-moment chair system (MMCS). It presents the overall design and changes that were implemented to compensate for problems. These were primarily to improve the accuracy of hip joint moments; a compromise between accuracy and practicalities, because of force-moment responses being measured at the ankles. All joint moment errors have been calculated to be of the order of a few newton metres. Since these represent errors of less than 10% when considering peak joint moment responses, this is considered sufficiently accurate for the proposed application. The MMCS is being used in a programme to investigate the restoration of lower limb functions, principally standing, in paraplegics by electrical stimulation of the lumbosacral anterior roots.     

基于微机电系统加速度传感器的上肢运动功能评价系统

目的设计一种基于微机电系统(micro-electro-mechical system,MEMS)加速度传感器的上肢运动功能评价系统,通过平滑度和加速度均方根指标精确客观地评价脑卒中患者上肢运动的功能障碍。方法本系统主要包括ATmega128单片机的通信模块,可进行上下位机的数据传输,保存数据;ADXL345加速度传感器的数据采集模块,实时采集患者的上肢运动加速度,本系统造价低、便于携带、测试高效。选择6例脑卒中上肢偏瘫康复患者(病患组)和4例上肢健康者(健康组)进行初步测试。健康组进行零负荷定向够物运动和梯度负荷够物运动重复测试;病患组和健康组进行零负荷定向够物运动对比测试。结果零负荷定向够物运动重复测试:健康组被试者间的平滑度(P0.05)和加速度均方根(P0.05)差异无统计学意义;零负荷定向够物运动的对比测试:患者与健康人员的平滑度指标差异具有统计学意义(P0.05),加速度均方根指标差异无统计学意义(P0.05);梯度负荷够物运动重复测试:平滑度随着负荷的增加而增长,且经一元线性回归分析,平滑度与负荷线性关系显著(P0.001)。结论基于MEMS加速度传感器的上肢运动功能评价系统可通过平滑度定量评价脑卒中患者的上肢运动能力,并进一步得到平滑度和负荷之间呈线性关系,为后续评价脑卒中上肢偏瘫患者的康复状况,提供了新的思路。     

A new shoulder model with a biologically inspired glenohumeral joint

Kinematically unconstrained biomechanical models of the glenohumeral (GH) joint are needed to study the GH joint function, especially the mechanisms of joint stability. The purpose of this study is to develop a large-scale multibody model of the upper limb that simulates the 6 degrees of freedom (DOF) of the GH joint and to propose a novel inverse dynamics procedure that allows the evaluation of not only the muscle and joint reaction forces of the upper limb but also the GH joint translations. The biomechanical model developed is composed of 7 rigid bodies, constrained by 6 anatomical joints, and acted upon by 21 muscles. The GH joint is described as a spherical joint with clearance. Assuming that the GH joint translates according to the muscle load distribution, the redundant muscle load sharing problem is formulated considering as design variables the 3 translational coordinates associated with the GH joint translations, the joint reaction forces associated with the remaining kinematic constraints, and the muscle activations. For the abduction motion in the frontal plane analysed, the muscle and joint reaction forces estimated by the new biomechanical model proposed are similar to those estimated by a model in which the GH joint is modeled as an ideal spherical joint. Even though this result supports the assumption of an ideal GH joint to study the muscle load sharing problem, only a 6 DOF model of the GH joint, as the one proposed here, provides information regarding the joint translations. In this study, the biomechanical model developed predicts an initial upward and posterior migration of the humeral head, followed by an inferior and anterior movement, which is in good agreement with the literature.     

Measurement of upper extremity orientation by video stereometry system

In the attempt to gain a broader understanding of the causal relationships behind work-related symptoms of pain in the human shoulder, monitoring of arm position is crucial. Different methods have been used with varying accuracy. A video-based stereometry system, using infra-red light and reflecting markers for motion analysis, has been introduced for measurements in the fields of ergonomics, biomechanics and sports medicine. The purpose of this study is to investigate the sources of error in using this system for posture registration of the upper limb. Measurements are performed on a calibration fixture, on a mechanical model of the upper limb and on a subject with an exoskeleton. Particular, attention is given to inconsistencies and relative errors due to the finite geometrical precision with which the markers are positioned in the calibration fixture and on the studied objects, the limited capability to align the objects relative to the coordinate system of the calibration fixture and the errors connected to angular measurements using protractors etc. It is concluded that the system makes a valuable addition to existing instruments for non-contact posture measurement, and produces position data with an adequate accuracy in normal handling.     

Contrasting roles of inertial and muscle moments at knee and ankle during paw-shake response

Intralimb kinetics of the paw-shake response (PSR) were studied in four spinal, adult cats. Using rigid body equations of motion to determine the dynamic interactions between limb segments, knee and ankle joint kinetics were calculated for the steady-state cycles as defined in the preceding paper. Hindlimb motion was filmed (200 frames/s) to obtain knee and ankle kinematics. Responses of flexors and extensors at both joints were recorded synchronously with cinefilm. Ankle and knee joint kinematics were determined from 51 steady-state cycles of 16 PSRs. Average maximum displacements, velocities, and accelerations were substantially greater for the ankle than for the knee joint. Knee and ankle motions were out of phase in the first part of the cycle; knee extension occurred simultaneously with ankle flexion. In the second part of the cycle, motions at the two joints were sequential; rapid knee flexion, accompanied by negligible ankle displacement, preceded rapid ankle extension with minimal knee displacement. At the ankle joint, peak net moments tending to cause flexion and extension were similar in magnitude and determined primarily by muscle moments. Moments due to leg angular acceleration contributed significantly to an extensor peak in the net moment near the end of the cycle. Other inertial and gravitational moments were small. At the knee joint, net moments tending to cause flexion and extension were also similar, but smaller than those at the ankle. The knee muscle moments, however, were large and counteracted large inertial moments due to paw angular acceleration. Also, moments due to leg angular acceleration and knee linear acceleration were substantial and opposite in effect. Other inertial and the gravitational moments were negligible. Muscle moments slowed and reversed joint motions, and active muscle force components of muscle moments were derived from lengthening of active musculotendinous units. Segmental interactions, in which proximal segment motion augmented distal segment velocity, increased the effectiveness of PSR steady-state cycles by facilitating the generation of extremely large paw linear accelerations. Limb oscillations during PSR steady-state result from interactions between muscle synergies and motion-dependent limb dynamics. At the ankle, muscle activity functioned to control paw acceleration, whereas at the knee, muscle activity functioned to control leg and paw inertial interactions.(ABSTRACT TRUNCATED AT 400 WORDS)     

基于肌电信号的膝关节跨越障碍角度预测方法

为解决人体跨越障碍物时膝关节角度输出的问题,针对性设计一种穿戴式信号获取实验台,对下肢运动姿态进行运动分析,将肌肉电信号及关节角度信号作为运动数据,对信号进行处理后利用BP神经网络预测跨越障碍时输出角度,提出一种利用BP神经网络算法,根据不同大腿抬起高度,分析膝关节运动主动肌与被动肌发力程度,预测输出人体跨越障碍时膝关节角度的方法,能够有效帮助假肢膝关节或康复机器人实现跨越障碍的复杂动作。     

A Spherical Rotation Coordinate System for the Description of Three-Dimensional Joint Rotations

Three-dimensional joint rotations in human movement analysis have been mainly described by Euler/Cardan angles. Due to sequence dependence, each combination of three Euler/Cardan angles defines a single pattern of joint rotation. When the rotation pattern is unknown, it needs to be considered using a particular sequence of Euler/Cardan angles to represent joint rotations. In this paper a spherical rotation coordinate system is developed for describing three-dimensional joint rotations using a method of rotation involving two steps: a long axis rotation and a pure axial rotation. Two angles of the classical spherical coordinate system--longitude and latitude--are used to describe long axis rotations in this newly proposed coordinate system. The spherical rotation coordinate system uses a radial rotation angle to describe pure axial rotation of a limb segment whereas the classical spherical coordinate system uses a radial displacement to describe motion of a point. An application of the spherical rotation coordinate system is given to define three-dimensional rotations of the glenohumeral joint. A mathematical proof shows that the long axis rotation and axial rotation are sequence independent. Two numerical examples are investigated which demonstrate that the spherical rotation angles can be uniquely determined in both forward and inverse kinematics without considering sequences rotations.