Objective. To investigate whether a microelectromechanical system (MEMS) inertial sensor module is as accurate as fiber-optic gyroscopes when classifying subjects as normal for clinical stance and gait balance tasks. Methods. Data of ten healthy subjects were recorded simultaneously with a fiber-optic gyroscope (FOG) system of SwayStar™ and a MEMS sensor system incorporated in the Valedo® system. Data from a sequence of clinical balance tasks with different angle and angular velocity ranges were assessed. Paired t-tests were performed to determine significant differences between measurement systems. Cohen’s kappa test was used to determine the classification of normal balance control between the two sensor systems when comparing the results to a reference database recorded with the FOG system. Potential cross-talk errors in roll and pitch angles when neglecting yaw axis rotations were evaluated by comparing 2D FOG and 3D MEMS recordings. Results. Statistically significant (α=0.05) differences were found in some balance tasks, for example, “walking eight tandem steps” and various angular measures (p<0.03). However, these differences were within a few percent (<2.7%) of the reference values. Tasks with high dynamic velocity ranges showed significant differences (p=0.002) between 2D FOG and 3D MEMS roll angles but no difference between 2D FOG and 2D MEMS roll angles. An almost perfect agreement could be obtained for both 2D FOG and 2D MEMS (κ=0.97) and 2D FOG and 3D MEMS measures (κ=0.87) when comparing measurements of all subjects and tasks. Conclusion. MEMS motion sensors can be used for assessing balance during clinical stance and gait tasks. MEMS provides measurements comparable to values obtained with a highly accurate FOG. When assessing pitch and roll trunk sway measures without accounting for the effect of yaw, it is recommended to use angle and angular velocity measures for stance, and only angular velocity measures for gait because roll and pitch velocity measurements are not influenced by yaw rotations, and angle errors are low for stance.

Introduction and past research: Inertial sensors are great for orientation estimation, but they cannot measure relative positions of human body segments directly. In previous work we used ultrasound to estimate distances between body segments [1]. In [2] we presented an easy to use system for gait analysis in clinical practice but also in-home situations. Ultrasound range estimates were fused with data from foot-mounted inertial sensors, using an extended Kalman filter, for 3D (relative) position and orientation estimation of the feet. Validation: From estimated 3D positions we calculated step lengths and stride widths and compared this to an optical reference system for validation. Mean (±standard deviation) of absolute differences was 1.7 cm (±1.8 cm) for step lengths and 1.2 cm (±1.2 cm) for stride widths when comparing 54 walking trials of three healthy subjects. Clinical application: Next, the system presented in [2] was used in the INTERACTION project, for measuring eight stroke subjects during a 10 m walk test [3]. Step lengths, stride widths and stance and swing times were compared with the Berg balance scale score. The first results showed a correlation between step lengths and Berg balance scale scores. To draw real conclusions, more patients and also different activities will be investigated next. Future work: In future work we will extend the system with inertial sensors on the upperand lower legs and the pelvis, to be able to calculate a closed loop and improve the estimation of joint angles compared with systems containing only inertial sensors.

Relative foot position estimation is important for rehabilitation, sports training and functional diagnostics. In this paper an extended Kalman filter fusing ultrasound range estimates and inertial sensors is described. With this filter several gait parameters can be estimated ambulatory. Step lengths and stride widths from 54 walking trials of three healthy subjects were estimated and compared to an optical reference. Mean ( ± standard deviation) of absolute difference was 1.7 cm ( ±1.8 cm) and 1.2 cm ( ± 1.2 cm) for step length and stride width respectively. Walking with a turn and walking around in a square area were also investigated and resulted in mean absolute differences of 1.7 cm ( ±2.0 cm) and 1.5 cm ( ± 1.5 cm) for step lengths and stride widths. In addition to these relative positions, velocities, orientations and stance and swing times can also be estimated. We conclude that the presented system is low-cost and provides a complete description of footstep kinematics and timing.

BACKGROUND:Assessment of hand kinematics is important when evaluating hand functioning. Major drawbacks of current sensing glove systems are lack of rotational observability in particular directions, labour intensive calibration methods which are sensitive to wear and lack of an absolute hand orientation estimate.METHODS:We propose an ambulatory system using inertial sensors that can be placed on the hand, fingers and thumb. It allows a full 3D reconstruction of all finger and thumb joints as well as the absolute orientation of the hand. The system was experimentally evaluated for the static accuracy, dynamic range and repeatability.RESULTS:The RMS position norm difference of the fingertip compared to an optical system was 5±0.5 mm (mean ± standard deviation) for flexion-extension and 12.4±3.0 mm for combined flexion-extension abduction-adduction movements of the index finger. The difference between index and thumb tips during a pinching movement was 6.5±2.1 mm. The dynamic range of the sensing system and filter was adequate to reconstruct full 80 degrees movements of the index finger performed at 116 times per minute, which was limited by the range of the gyroscope. Finally, the reliability study showed a mean range difference over five subjects of 1.1±0.4 degrees for a flat hand test and 1.8±0.6 degrees for a plastic mold clenching test, which is smaller than other reported data gloves.CONCLUSION:Compared to existing data gloves, this research showed that inertial and magnetic sensors are of interest for ambulatory analysis of the human hand and finger kinematics in terms of static accuracy, dynamic range and repeatability. It allows for estimation of multi-degree of freedom joint movements using low-cost sensors.

To support real-time, drift-free human motion capture of multiple closely interacting actors, Xsens introduces a revolutionary product: MVN Motiongrid. This system is based on tightly coupled ultra-wideband radio based positioning and inertial motion capture, and it is available as an optional add-on to the Xsens MVN full body motion capture system. This combination is capable of eliminating the incremental drift inevitable in motion capture based on inertial sensors alone, without affecting the unique advantages of MVN, namely immunity to occlusions, cost effective large area capture, and flexibility of use. This paper describes the basic working principles, architecture, and performance of Xsens MVN MotionGrid.

Over the last years, inertial sensing has proven to be a suitable ambulatory alternative to traditional human motion tracking based on optical position measurement systems, which are generally restricted to a laboratory environment. Besides many advantages, a major drawback is the inherent drift caused by integration of acceleration and angular velocity to obtain position and orientation. In addition, inertial sensing cannot be used to estimate relative positions and orientations of sensors with respect to each other. In order to overcome these drawbacks, this study presents an Extended Kalman Filter for fusion of inertial and magnetic sensing that is used to estimate relative positions and orientations. In between magnetic updates, change of position and orientation are estimated using inertial sensors. The system decides to perform a magnetic update only if the estimated uncertainty associated with the relative position and orientation exceeds a predefined threshold. The filter is able to provide a stable and accurate estimation of relative position and orientation for several types of movements, as indicated by the average rms error being 0.033 m for the position and 3.6 degrees for the orientation.

The Xsens MVN motion capture suit is an easy-to- use, cost efficient system for full-body human motion capture. MVN is based on unique, state-of-the-art miniature inertial sensors, biomechanical models and sensor fusion algorithms. MVN does not need external cameras, emitters or markers. It can thus be used outdoors as well as indoors, there are no restrictions for lighting, it does not suffer from problems of occlusion or missing markers. In addition, unique for inertial motion capture technology: the sensor-suit captures any type of movement, including running, jumping, crawling and cartwheels.

The use of miniature inertial sensors has become a common practice in ambulatory motion analysis. For accurate and drift free orientation estimation several methods have been reported combining the signals from 3D gyroscopes, accelerometers and magnetometers. Accelerometers are used to determine the direction of the local vertical by sensing acceleration due to gravity. Magnetic sensors provide stability in the horizontal plane by sensing the direction of the earth magnetic field like a compass. Data from these complementary sensors can be used to eliminate drift by continuous correction of the orientation obtained by integrating rate sensor data [1].

Inertial and magnetic sensors are very suitable for ambulatory monitoring of human posture and movements. However, ferromagnetic materials near the sensor disturb the local magnetic field and, therefore, the orientation estimation. A Kalman-based fusion algorithm was used to obtain dynamic orientations and to minimize the effect of magnetic disturbances. This paper compares the orientation output of the sensor fusion using three-dimensional inertial and magnetic sensors against a laboratory bound opto-kinetic system (Vicon) in a simulated work environment. With the tested methods, the difference between the optical reference system and the output of the algorithm was 2.6 degrees root mean square (rms) when no metal was near the sensor module. Near a large metal object instant errors up to 50 degrees were measured when no compensation was applied. Using a magnetic disturbance model, the error reduced significantly to 3.6 degrees rms.

This paper presents the design and testing of a portable magnetic system combined with miniature inertial sensors for ambulatory 6 degrees of freedom (DOF) human motion tracking. The magnetic system consists of three orthogonal coils, the source, fixed to the body and 3-D magnetic sensors, fixed to remote body segments, which measure the fields generated by the source. Based on the measured signals, a processor calculates the relative positions and orientations between source and sensor. Magnetic actuation requires a substantial amount of energy which limits the update rate with a set of batteries. Moreover, the magnetic field can easily be disturbed by ferromagnetic materials or other sources. Inertial sensors can be sampled at high rates, require only little energy and do not suffer from magnetic interferences. However, accelerometers and gyroscopes can only measure changes in position and orientation and suffer from integration drift. By combing measurements from both systems in a complementary Kalman filter structure, an optimal solution for position and orientation estimates is obtained. The magnetic system provides 6 DOF measurements at a relatively low update rate while the inertial sensors track the changes position and orientation in between the magnetic updates. The implemented system is tested against a lab-bound camera tracking system for several functional body movements. The accuracy was about 5 mm for position and 3 degrees for orientation measurements. Errors were higher during movements with high velocities due to relative movement between source and sensor within one cycle of magnetic actuation.

This paper presents the design and testing of a portable magnetic system for human motion tracking. Three essential components comprise this system (1) 3D source, consisting of three orthogonal coils, placed on the body; (2) a compatible 3D sensor, which is fixed at a remote body segment and measures the fields generated by the source; (3) a processor whose function is to relate the signals from source and sensor. Given the signals from the source and sensor, the position and orientation of the sensor in 6 degrees of freedom (DOF) with respect to the position of the transmitter can be estimated. Coil parameters, such as the radius are optimized for tracking the distance between the lower back and the shoulder of a person. The electronics are designed to run on battery supply, making it suitable for body mounting and ambulatory measurements. The system is tested with functional body movements such as flexion of the back, arm movement and walking. The accuracy of measurements is approximately 8mm in position and 5° in orientation with 6 DOF movements. Results are less accurate during relatively high velocities of the source or sensor due to under sampling. The magnetic tracker will be used as an aiding system for fusion with inertial sensors, therefore, the update rate requirements can be relatively low.

A motion tracking system for tracking an object composed of object parts in a three-dimensional space. The system comprises a number of magnetic field transmitters; a number of field receivers for receiving the magnetic fields of the field transmitters; a number of inertial measurement units for recording a linear acceleration; a number of angular velocity transducers for recording angular velocities. The system further comprises a processor for controlling the transmitters and receiving signals coming from the field receivers and the inertial measurement unit; which processor contains a module for deriving orientation and/or position information of the constituent object parts of the object on the basis of the received signals. The processor is configured for intermittently controlling the transmitters transmit at a predetermined frequency, wherein the position and/or orientation information is derived by periodically calibrating the motion information coming from the inertial measurement unit with the motion information coming from the magnetic field receivers.

This paper describes a complementary Kalman filter design to estimate orientation of human body segments by fusing gyroscope, accelerometer, and magnetometer signals from miniature sensors. Ferromagnetic materials or other magnetic fields near the sensor module disturb the local earth magnetic field and, therefore, the orientation estimation, which impedes many (ambulatory) applications. In the filter, the gyroscope bias error, orientation error, and magnetic disturbance error are estimated. The filter was tested under quasi-static and dynamic conditions with ferromagnetic materials close to the sensor module. The quasi-static experiments implied static positions and rotations around the three axes. In the dynamic experiments, three-dimensional rotations were performed near a metal tool case. The orientation estimated by the filter was compared with the orientation obtained with an optical reference system Vicon. Results show accurate and drift-free orientation estimates. The compensation results in a significant difference (p < 0.01) between the orientation estimates with compensation of magnetic disturbances in comparison to no compensation or only gyroscopes. The average static error was 1.4 degrees (standard deviation 0.4) in the magnetically disturbed experiments. The dynamic error was 2.6 degrees root means square.

A method for combining a camera-marker based motion analysis system with miniature inertial sensors is proposed. It is used to fill gaps of optical data and can increase the data rate of the optical system.

Small inertial sensors like accelerometers and gyroscopes are more and more used in ambulatory motion analysis (Busseral. 1998; Baten et al. 2000; Veltink et al. 2003). Typically, angular orientation of a body segment is determined by integrating the output from the angular rate sensors strapped on the segment.