Gait analysis using gravitational acceleration measured by wearable sensors

A novel method for measuring human gait posture using wearable sensor units is proposed. The sensor units consist of a tri-axial acceleration sensor and three gyro sensors aligned on three axes. The acceleration and angular velocity during walking were measured with seven sensor units worn on the abdomen and the lower limb segments (both thighs, shanks and feet). The three-dimensional positions of each joint are calculated from each segment length and joint angle. Joint angle can be estimated mechanically from the gravitational acceleration along the anterior axis of the segment. However, the acceleration data during walking includes three major components; translational acceleration, gravitational acceleration and external noise. Therefore, an optimization analysis was represented to separate only the gravitational acceleration from the acceleration data. Because the cyclic patterns of acceleration data can be found during constant walking, a FFT analysis was applied to obtain some characteristic frequencies in it. A pattern of gravitational acceleration was assumed using some parts of these characteristic frequencies. Every joint position was calculated from the pattern under the condition of physiological motion range of each joint. An optimized pattern of the gravitational acceleration was selected as a solution of an inverse problem. Gaits of three healthy volunteers were measured by walking for 20 s on a flat floor. As a result, the acceleration data of every segment was measured simultaneously. The characteristic three-dimensional walking could be shown by the expression using a stick figure model. In addition, the trajectories of the knee joint in the horizontal plane could be checked by visual imaging on a PC. Therefore, this method provides important quantitive information for gait diagnosis.     

Gait and posture discrimination in sheep using a tri-axial accelerometer

Temporo-spatial observation of the leg could provide important information about the general condition of an animal, especially for those such as sheep and other free-ranging farm animals that can be difficult to access. Tri-axial accelerometers are capable of collecting vast amounts of data for locomotion and posture observations; however, interpretation and optimization of these data records remain a challenge. The aim of the present study was to introduce an optimized method for gait (walking, trotting and galloping) and posture (standing and lying) discrimination, using the acceleration values recorded by a tri-axial accelerometer mounted on the hind leg of sheep. The acceleration values recorded on the vertical and horizontal axes, as well as the total acceleration values were categorized. The relative frequencies of the acceleration categories (RFACs) were calculated in 3-s epochs. Reliable RFACs for gait and posture discrimination were identified with discriminant function and canonical analyses. Post hoc predictions for the two axes and total acceleration were conducted, using classification functions and classification scores for each epoch. Mahalanobis distances were used to determine the level of accuracy of the method. The highest discriminatory power for gait discrimination yielded four RFACs on the vertical axis, and five RFACs each on the horizontal axis and total acceleration vector. Classification functions showed the highest accuracy for walking and galloping. The highest total accuracy on the vertical and horizontal axes were 90% and 91%, respectively. Regarding posture discrimination, the vertical axis exhibited the highest discriminatory power, with values of RFAC (0, 1]=99.95% for standing; and RFAC (−1, 0]=99.50% for lying. The horizontal axis showed strong discrimination for the lying side of the animal, as values were in the acceleration category of (0, 1] for lying on the left side and (−1, 0] on the right side. The algorithm developed by the method employed in the present study facilitates differentiation of the various types of gait and posture in animals from fewer data records, and produces the most reliable acceleration values from only one axis within a short time frame. The present study introduces an optimized method by which the tri-axial accelerometer can be used in gait and posture discrimination in sheep as an animal model.     

A fast estimation of biochemical oxygen demand using microbial sensors

Summary Microbial amperometric sensors for biochemical oxygen demand (BOD) determination using Bacillus subtilis or Trichosporon cutaneum cells immobilized in polyvinylalcohol have been developed. These sensors allow BOD measurements with very short response times (15–30s), a level of precision of ±5% and an operation stability of 30 days. A linear range was obtained for a B. subtilis-based sensor up to 20 mg/l BOD and for a T. cutaneum-based sensor up to 100 mg/l BOD using a glucose/glutamic acid standard.     

Ambulatory estimation of foot placement during walking using inertial sensors

This study proposes a method to assess foot placement during walking using an ambulatory measurement system consisting of orthopaedic sandals equipped with force/moment sensors and inertial sensors (accelerometers and gyroscopes). Two parameters, lateral foot placement (LFP) and stride length (SL), were estimated for each foot separately during walking with eyes open (EO), and with eyes closed (EC) to analyze if the ambulatory system was able to discriminate between different walking conditions. For validation, the ambulatory measurement system was compared to a reference optical position measurement system (Optotrak). LFP and SL were obtained by integration of inertial sensor signals. To reduce the drift caused by integration, LFP and SL were defined with respect to an average walking path using a predefined number of strides. By varying this number of strides, it was shown that LFP and SL could be best estimated using three consecutive strides. LFP and SL estimated from the instrumented shoe signals and with the reference system showed good correspondence as indicated by the RMS difference between both measurement systems being 6.5±1.0 mm (mean ±standard deviation) for LFP, and 34.1±2.7 mm for SL. Additionally, a statistical analysis revealed that the ambulatory system was able to discriminate between the EO and EC condition, like the reference system. It is concluded that the ambulatory measurement system was able to reliably estimate foot placement during walking.     

Bottom-up estimation of joint moments during manual lifting using orientation sensors instead of position sensors

L5/S1, hip and knee moments during manual lifting tasks are, in a laboratory environment, frequently established by bottom-up inverse dynamics, using force plates to measure ground reaction forces (GRFs) and an optoelectronic system to measure segment positions and orientations. For field measurements, alternative measurement systems are being developed. One alternative is the use of small body-mounted inertial/magnetic sensors (IMSs) and instrumented force shoes to measure segment orientation and GRFs, respectively. However, because IMSs measure segment orientations only, the positions of segments relative to each other and relative to the GRFs have to be determined by linking them, assuming fixed segment lengths and zero joint translation. This will affect the estimated joint positions and joint moments. This study investigated the effect of using segment orientations only (orientation-based method) instead of using orientations and positions (reference method) on three-dimensional joint moments. To compare analysis methods (and not measurement methods), GRFs were measured with a force plate and segment positions and/or orientations were measured using optoelectronic marker clusters for both analysis methods. Eleven male subjects lifted a box from floor level using three lifting techniques: a stoop, a semi-squat and a squat technique. The difference between the two analysis methods remained small for the knee moments: <4%. For the hip and L5/S1 moments, the differences were more substantial: up to 8% for the stoop and semi-squat techniques and up to 14% for the squat technique. In conclusion, joint moments during lifting can be estimated with good accuracy at the knee joint and with reasonable accuracy at the hip and L5/S1 joints using segment orientation and GRF data only.     

Validation of a model-based inverse kinematics approach based on wearable inertial sensors

Abstract

Wearable inertial measurement units (IMUs) are a promising solution to human motion estimation. Using IMUs 3D orientations, a model-driven inverse kinematics methodology to estimate joint angles is presented. Estimated joint angles were validated against encoder-measured kinematics (robot) and against marker-based kinematics (passive mechanism). Results are promising, with RMS angular errors respectively lower than 3 and 6?deg over a minimum range of motion of 50?deg (robot) and 160?deg (passive mechanism). Moreover, a noise robustness analysis revealed that the model-driven approach reduces the effects of experimental noises, making the proposed technique particularly suitable for application in human motion analysis.     

Combining position and acceleration measurements for joint force estimation

The calculation of joint forces in biomechanics is usually based on the measurements of the kinematics of a given body segment, the estimation of the inertial properties of that segment and the solution of the ‘inverse dynamics problem’. Such a process results in estimates of the joint forces and moments needed to sustain the monitored motion. This paper presents a new approach that combines position and acceleration measurements for the purpose of deriving high-quality joint force estimates. An experimental system that is based on an instrumented compound pendulum was designed and tested. The joint forces necessary to maintain a swinging motion of the pendulum were measured by an array of strain gauges, and were compared to the forces estimated by the integrated kinematic segment that measured the position and acceleration of the pendulum. The joint force measurements were also compared to the force estimates that were based on the calculated segmental acceleration generated by the differentiation of the segmental position alone. The results show a high degree of correlation between the forces estimated by the integrated segment and those measured by the strain gauges. The force estimates based on the position measurements alone were less accurate and noisier. The application of the integrated segment to the study of human kinetics is discussed and illustrated by the ankle and knee forces during slow walking. The results suggest that the use of accelerometers is necessary for the estimation of transients and high-frequency components of joint forces.     

Egomotion estimation with optic flow and air velocity sensors

We develop a method that allows a flyer to estimate its own motion (egomotion), the wind velocity, ground slope, and flight height using only inputs from onboard optic flow and air velocity sensors. Our artificial algorithm demonstrates how it could be possible for flying insects to determine their absolute egomotion using their available sensors, namely their eyes and wind sensitive hairs and antennae. Although many behaviors can be performed by only knowing the direction of travel, behavioral experiments indicate that odor tracking insects are able to estimate the wind direction and control their absolute egomotion (i.e., groundspeed). The egomotion estimation method that we have developed, which we call the opto-aeronautic algorithm, is tested in a variety of wind and ground slope conditions using a video recorded flight of a moth tracking a pheromone plume. Over all test cases that we examined, the algorithm achieved a mean absolute error in height of 7% or less. Furthermore, our algorithm is suitable for the navigation of aerial vehicles in environments where signals from the Global Positioning System are unavailable.     

Close encounters in a pediatric ward: measuring face-to-face proximity and mixing patterns with wearable sensors

Background

Nosocomial infections place a substantial burden on health care systems and represent one of the major issues in current public health, requiring notable efforts for its prevention. Understanding the dynamics of infection transmission in a hospital setting is essential for tailoring interventions and predicting the spread among individuals. Mathematical models need to be informed with accurate data on contacts among individuals.

Methods and Findings

We used wearable active Radio-Frequency Identification Devices (RFID) to detect face-to-face contacts among individuals with a spatial resolution of about 1.5 meters, and a time resolution of 20 seconds. The study was conducted in a general pediatrics hospital ward, during a one-week period, and included 119 participants, with 51 health care workers, 37 patients, and 31 caregivers. Nearly 16,000 contacts were recorded during the study period, with a median of approximately 20 contacts per participants per day. Overall, 25% of the contacts involved a ward assistant, 23% a nurse, 22% a patient, 22% a caregiver, and 8% a physician. The majority of contacts were of brief duration, but long and frequent contacts especially between patients and caregivers were also found. In the setting under study, caregivers do not represent a significant potential for infection spread to a large number of individuals, as their interactions mainly involve the corresponding patient. Nurses would deserve priority in prevention strategies due to their central role in the potential propagation paths of infections.

Conclusions

Our study shows the feasibility of accurate and reproducible measures of the pattern of contacts in a hospital setting. The obtained results are particularly useful for the study of the spread of respiratory infections, for monitoring critical patterns, and for setting up tailored prevention strategies. Proximity-sensing technology should be considered as a valuable tool for measuring such patterns and evaluating nosocomial prevention strategies in specific settings.     

Analyzing wearable device data using marked point processes

This paper introduces two sets of measures as exploratory tools to study physical activity patterns: active‐to‐sedentary/sedentary‐to‐active rate function (ASRF/SARF) and active/sedentary rate function (ARF/SRF). These two sets of measures are complementary to each other and can be effectively used together to understand physical activity patterns. The specific features are illustrated by an analysis of wearable device data from National Health and Nutrition Examination Survey (NHANES). A two‐level semiparametric regression model for ARF and the associated activity magnitude is developed under a unified framework using the marked point process formulation. The inactive and active states measured by accelerometers are treated as a 0‐1 point process, and the activity magnitude measured at each active state is defined as a marked variable. The commonly encountered missing data problem due to device nonwear is referred to as “window censoring,” which is handled by a proper estimation approach that adopts techniques from recurrent event data. Large sample properties of the estimator and comparison between two regression models as measurement frequency increases are studied. Simulation and NHANES data analysis results are presented. The statistical inference and analysis results suggest that ASRF/SARF and ARF/SRF provide useful analytical tools to practitioners for future research on wearable device data.     

New considerations for collecting biomechanical data using wearable sensors: Number of level runs to define a stable running pattern with a single IMU

Wearable technology can be used to quantify running biomechanical patterns in a runner’s natural environment, however, changes in external factors during outdoor running may influence a runner’s typical gait pattern. Therefore, the purpose of this study was to determine how many runs are needed to define a stable or typical running pattern. Six biomechanical variables were recorded using a single wearable sensor placed on the lower back during ten outdoor runs for twelve runners. Univariate and multivariate distributions were created and based on the probability density function, the percent of similar data points (within 95%) from each unique run for the same runner were determined. Stability was defined when the addition of data from a new run resulted in less than a 5% change in the probability density function. To cross-validate, the percent of similar data points at stability was compared between the same and different runners using a repeated-measures MANOVA (Bonferroni-corrected α = 0.007). The maximum number of runs needed to reach stability for univariate and multivariate analyses was four and five, respectively. There was a significant overall effect on similar data points between the same and different runners (p = 0.001), with a greater percent of similar data points for the same runner compared to other runners (p < 0.007). Based on biomechanical data collected using a single wearable sensor placed on the lower back, this is the first study to show that four (univariate) to five (multivariate) runs are needed to establish a stable running pattern in real-world settings.     

Sleeping posture recognition using fuzzy c-means algorithm

Background

Pressure sensors have been used for sleeping posture detection, which meet privacy requirements. Most of the existing techniques for sleeping posture recognition used force-sensitive resistor (FSR) sensors. However, lower limbs cannot be recognized accurately unless thousands of sensors are deployed on the bedsheet.

Method

We designed a sleeping posture recognition scheme in which FSR sensors were deployed on the upper part of the bedsheet to record the pressure distribution of the upper body. In addition, an infrared array sensor was deployed to collect data for the lower body. Posture recognition was performed using a fuzzy c-means clustering algorithm. Six types of sleeping body posture were recognized from the combination of the upper and lower body postures.

Results

The experimental results showed that the proposed method achieved an accuracy of above 88%. Moreover, the proposed scheme is cost-efficient and easy to deploy.

Conclusions

The proposed sleeping posture recognition system can be used for pressure ulcer prevention and sleep quality assessment. Compared to wearable sensors and cameras, FSR sensors and infrared array sensors are unobstructed and meet privacy requirements. Moreover, the proposed method provides a cost-effective solution for the recognition of sleeping posture.

     

Studies on spatial orientation and posture control and changes in otolith function due to linear acceleration loading.

Otolith function is directly affected by weightlessness at the time of movement in outer space, and changes occur in the mode of response. It has been known for some time that such changes occur in the posture and gait of astronauts just after they return from a trip into space. It is thought that the cause of these changes is disuse atrophy of the antigravity muscles. However, in the present study, experimental subjects underwent repeated linear acceleration loading over a long period of time, and instability of the head and a decrease in posture control, especially in relation to the gait, were observed for the first time. To date, it has been said that the otolith function has a close relationship with ocular counter rolling. However, when the otolith organ was stimulated, the response was seen to be head instability and an irregular effect on the gait. It is surmised that these findings will facilitate future research into the otolith function under gravity-free conditions.     

Subject-specific and group-based running pattern classification using a single wearable sensor

The objective of this study was to determine whether subject-specific or group-based models provided better classification accuracy to identify changes in biomechanical running gait patterns across different inclination conditions. The classification process was based on measurements from a single wearable sensor using a total of 41,780 strides from eleven recreational runners while running in real-world and uncontrolled environment. Biomechanical variables included pelvic drop, ground contact time, braking, vertical oscillation of pelvis, pelvic rotation, and cadence were recorded during running on three inclination grades: downhill, −2° to −7°; level, −0.2° to +0.2°; and uphill, +2° to +7°. An ensemble and non-linear machine learning algorithm, random forest (RF), was used to classify inclination condition and determine the importance of each of the biomechanical variables. Classification accuracy was determined for subject-specific and group-based RF models. The mean classification accuracy of all subject-specific RF models was 86.29%, while group-based classification accuracy was 76.17%. Braking was identified as the most important variable for all the runners using the group-based model and for most of the runners based on a subject-specific models. In addition, individual runners used different strategies across different inclination conditions and the ranked order of variable importance was unique for each runner. These results demonstrate that subject-specific models can better characterize changes in gait biomechanical patterns compared to a more traditional group-based approach.     

Measurement of stride parameters using a wearable GPS and inertial measurement unit

Both GPS and inertial measurement units (IMUs) have been extensively used in biomechanical studies. Expensive high accuracy GPS units can provide information about intrastride speed and position, but their application is limited by their size and cost. Single and double integration of acceleration from IMU provides information about short-term fluctuations in speed and position, but suffers from integration error over a longer period of time. The integration of GPS and IMU has been widely used in large and expensive units designed for survey and vehicle navigation. Here we propose a data fusion scheme, which is a Kalman filter based complementary filter and enhances the frequency response of the GPS and IMU used alone. We also report the design of a small (28 g) low cost GPS/IMU unit. Its accuracy after post-processing with the proposed data fusion scheme for determining average speed and intrastride variation was compared to a traditional high cost survey GPS. The low cost unit achieved an accuracy of 0.15 ms⁻¹ (s.d.) for horizontal speed in cycling and human running across a speed range of 3–10 ms⁻¹. The stride frequency and vertical displacement calculated based on measurements from the low cost GPS/IMU units had an s.d. of 0.08 Hz and 0.02 m respectively, compared to measurements from high performance OEM4 GPS units.     

Prediction of fingers posture using artificial neural networks

Predicting the hand and fingers posture during grasping tasks is an important issue in the frame of biomechanics. In this paper, a technique based on neural networks is proposed to learn the inverse kinematics mapping between the fingertip 3D position and the corresponding joint angles. Finger movements are obtained by an instrumented glove and are mapped to a multichain model of the hand. From the fingertip desired position, the neural networks allow predicting the corresponding finger joint angles keeping the specific subject coordination patterns. Two sets of movements are considered in this study. The first one, the training set, consisting of free fingers movements is used to construct the mapping between fingertip position and joint angles. The second one, constructed for testing purposes, is composed of a sequence of grasping tasks of everyday-life objects. The maximal mean error between fingertip measured position and fingertip position obtained from simulated joint angles and forward kinematics is 0.99+/-0.76mm for the training set and 1.49+/-1.62mm for the test set. Also, the maximal RMS error of joint angles prediction is 2.85 degrees and 5.10 degrees for the training and test sets respectively, while the maximal mean joint angles prediction error is -0.11+/-4.34 degrees and -2.52+/-6.71 degrees for the training and test sets, respectively. Results relative to the learning and generalization capabilities of this architecture are also presented and discussed.     

Physical sensor difference-based method and virtual sensor difference-based method for visual and quantitative estimation of lower limb 3D gait posture using accelerometers and magnetometers

An approach using a physical sensor difference-based algorithm and a virtual sensor difference-based algorithm to visually and quantitatively confirm lower limb posture was proposed. Three accelerometers and two MAG(3)s (inertial sensor module) were used to measure the accelerations and magnetic field data for the calculation of flexion/extension (FE) and abduction/adduction (AA) angles of hip joint and FE, AA and internal/external rotation (IE) angles of knee joint; then, the trajectories of knee and ankle joints were obtained with the joint angles and segment lengths. There was no integration of acceleration or angular velocity for the joint rotations and positions, which is an improvement on the previous method in recent literature. Compared with the camera motion capture system, the correlation coefficients in five trials were above 0.91 and 0.92 for the hip FE and AA, respectively, and higher than 0.94, 0.93 and 0.93 for the knee joint FE, AA and IE, respectively.     

A computer vision based method for 3D posture estimation of symmetrical lifting

Work-related musculoskeletal disorders (WMSD) are commonly observed among the workers involved in material handling tasks such as lifting. To improve work place safety, it is necessary to assess musculoskeletal and biomechanical risk exposures associated with these tasks. Such an assessment has been mainly conducted using surface marker-based methods, which is time consuming and tedious. During the past decade, computer vision based pose estimation techniques have gained an increasing interest and may be a viable alternative for surface marker-based human movement analysis. The aim of this study is to develop and validate a computer vision based marker-less motion capture method to assess 3D joint kinematics of lifting tasks. Twelve subjects performing three types of symmetrical lifting tasks were filmed from two views using optical cameras. The joints kinematics were calculated by the proposed computer vision based motion capture method as well as a surface marker-based motion capture method. The joint kinematics estimated from the computer vision based method were practically comparable to the joint kinematics obtained by the surface marker-based method. The mean and standard deviation of the difference between the joint angles estimated by the computer vision based method and these obtained by the surface marker-based method was 2.31 ± 4.00°. One potential application of the proposed computer vision based marker-less method is to noninvasively assess 3D joint kinematics of industrial tasks such as lifting.     

Physical sensor difference-based method and virtual sensor difference-based method for visual and quantitative estimation of lower limb 3D gait posture using accelerometers and magnetometers

An approach using a physical sensor difference-based algorithm and a virtual sensor difference-based algorithm to visually and quantitatively confirm lower limb posture was proposed. Three accelerometers and two MAG³s (inertial sensor module) were used to measure the accelerations and magnetic field data for the calculation of flexion/extension (FE) and abduction/adduction (AA) angles of hip joint and FE, AA and internal/external rotation (IE) angles of knee joint; then, the trajectories of knee and ankle joints were obtained with the joint angles and segment lengths. There was no integration of acceleration or angular velocity for the joint rotations and positions, which is an improvement on the previous method in recent literature. Compared with the camera motion capture system, the correlation coefficients in five trials were above 0.91 and 0.92 for the hip FE and AA, respectively, and higher than 0.94, 0.93 and 0.93 for the knee joint FE, AA and IE, respectively.