Adaptive neural fuzzy reasoning method for recognizing human movement gait phase

The study focuses on the fabrication of shoe sensors (MSS) that provides important information about the dynamic movement of the lower limbs. In this study the kinematic model of human lower limbs was analyzed to know and detect the changes during human gait phase, that occurs when walking, based on the GRF. The sensors number, sensor type and location of sensors was chosen to proceed on the MSS device. The algorithm was developed based on COP that is able to detect the gait phase by means of three sensors. The standard deviation and mean for COPy and vGRF values are expressed on the basis of walk of the five testees. It includes 4 gait phases: initial contact IC (1), mid-stance MS (2), terminal stance TS (3), and swing (4), the gait cycle detection is expressed as an average value. The experimental work is conducted to estimate the manufactured shoe sensor device and to detect the gait phase algorithm and compare it to an F-Socket sensor device (FSS). The results of the MSS device fabrication were (MS - mean value was 9.49 ± 2.76%, TS - mean value was 29.46 ± 4.95% and the swing mean value was 62.50 ± 1.60%) and the pressure result (255.629 KPа) while the result of the (FSS) device were (MS - mean value was 10.98 ± 1.74%, TS - mean value was 30.43 ± 2.32 %, and the swing mean value was 60.69 ± 1.74 %) and the pressure result (259.618 KPа) respectively, the test results showed that the results obtained from the (MSS) device are close to the readings obtained by using the (FSS) device, this shows the (MSS) device accuracy of this study despite the fact that it has simple design and inexpensive.

Background Gait analysis systems serve as important tools for assessing disturbed gait patterns. Amongst other factors, functional limitations of the shoulder joint may relate to such disturbances. Patient-reported outcome measures, assessment of pain, and active range of motion are commonly used to describe shoulder impairment. Purpose The aim of this cohort study was to evaluate the impact of unilateral limitations of shoulder mobility and pain on gait patterns and to detect correlations between pain, shoulder mobility, and particular phases of human gait using a Zebris gait analysis system. Methods 20 subjects with unilaterally restricted mobility and pain of the affected shoulder and a control group of 10 healthy subjects underwent a gait analysis. Various gait parameters, the DASH score, pain at rest and movement of the affected shoulder, and the active range of motion (aROM) for shoulder flexion and abduction were recorded. Results We determined significant differences of the duration of the loading response (p = 0.021), midstance (p = 0.033), and the terminal stance phase (p = 0.019) between the shoulder group and the control group, with a shorter loading response phase and a longer terminal stance phase of the affected side in the shoulder group. In the shoulder group, we found significant correlations between the DASH and the duration of the midstance phase (p = 0.023) and the terminal stance phase (p = 0.038). In addition, there was a significant correlation between shoulder flexion and the duration of the midstance phase (p = 0.047).

The human gait contains the information of lower limb movement posture. Gait recognition and analysis can provide control information for lower limb rehabilitation robots. In different gait phases, the position of the foot touching the ground is different. Therefore, gait classification can be performed according to plantar pressure. This study investigated the changes in plantar pressure information during human lower limb walking. Firstly, the STM32 single-chip microcomputer is used to collect the plantar pressure information detected by the high dynamic force sensing resistor(HD-FSR); secondly, the collected information is transmitted to the upper computer through Bluetooth; finally, all kinds of gait phases pressure features are analyzed and classified. Since the two constant parameters in the Support Vector Machine(SVM) classification algorithm have a great influence on the classification effect, a SVM classification algorithm based on the Particle Swarm Optimization(PSO) is proposed to realize the recognition of different phases.

Aiming at a pneumatic artificial muscle (PAM) lower extremity exoskeleton, a control mechanism based on hybrid phase sliding mode control (SMC) is proposed. First of all, the human gait cycle is mainly divided into the swing phase and stance phase, and the lower extremity exoskeleton phase models are established by the Euler–Lagrange method, respectively. Secondly, the lower limb exoskeleton is inevitably affected in the diverse working environment, and the exoskeleton model has nonlinear and strong coupling characteristics, which both increase the control difficulty. In this situations, a robust sliding mode control method is designed based on an Extended State Observer (ESO). Thirdly, the pneumatic muscle takes time to contract and relax, and then the joint input torque cannot jump when the gait phase changes, hence, the smoothing switching of the assistive control scheme is introduced to solve it. The smoothing switching time is detected by a phase detector, and the phase detector is designed by the plantar pressure information. Finally the comparative simulation shows that this control strategy has the advantages of fast time, high control precision and no jump during control torque switching. Pneumatic artificial muscle contraction rate curve shows that the pneumatic muscles’ motion range meets the control requirement of the exoskeleton.

Human gait analysis is a common but also challenging method to diagnose and characterize gait disorders. In recent years, wearable technologies have been investigated for gait analysis including gait phase detection with promising results. In this paper, we investigated the feasibility of a textile-based strain sensor integrated into an ankle brace to detect gait phases. This threadlike resistive sensor detects elongations and is, flexible and easy to integrate into textiles. In a laboratory study, five healthy subjects wearing the ankle brace were asked to walk at various low speeds while the sensor data were recorded. Three commonly used machine learning algorithms (random forest, support vector machines, and neural network) were investigated to differentiate between four phases (initial contact, mid-stance, pre-swing, swing) of the walking data. The random forest classifier performed best on our data set followed by neural network and support vector machines. Using to-fold cross-validation, an average accuracy of 95.49% ± 2.32% was achieved. This result compares well to other methods such as IMU-based technologies and demonstrates the feasibility of our approach of using textile-based strain sensors to detect gait phases in human walking.

Direct measurement of detailed kinematics of individual anatomical structures in the foot during human locomotion is crucial for understanding morphofunctional roles of the human foot structure that mechanically interacts with the ground in a favourable manner to maintain stable gait. In the present study, we constructed a dynamic gait simulator to load and mobilize the cadaver foot and directly measured the three-dimensional kinematics of foot bones using a biplane X-ray fluoroscopy. A robotic gait simulator was developed to load and mobilize the cadaver foot in a manner similar to the way human foot actually moves and interacts with the ground during walking. The simulator has three legs, fore, middle and hind legs, with the cadaver foot fixed to the middle leg (Figure ​(Figure1A).1A). After the middle cadaver foot contacts the ground, the hind leg departs from the ground. The fore leg replicates the foot-contact with the ground in the next step and toe-off of the middle cadaver foot follows afterward. Tendons of tibialis anterior and soleus were connected to pneumatic actuators to apply forces at the appropriate moment to reproduce how the foot would function during walking. Figure 1 A robotic gait simulator to load and mobilize the cadaver foot in a manner similar to the way human foot moves and interacts with the ground during walking (A). Automatic method to register bone surface models with the two fluoroscopic images recorded ... A biplanar dynamic X-ray fluoroscopic system was developed with Shimadzu Corporation, Kyoto, Japan. The system consists of two sets of x-ray sources and flat panels with a resolution of 2688 x 2208 pixels. We recorded biplanar X-ray videos of foot movement using the system using the dynamic cadaver model at 15 fps. For direct measurement of 3D kinematics of the foot bones, we developed an automatic method to register bone surface models with the two fluoroscopic images (Figure ​(Figure1B).1B). Specifically, the 3D surface models of the foot were generated based on computed tomography (CT), and a similarity measure between occluding contours of the bone surface models with edge-enhanced fluoroscopic images was evaluated to reconstruct the position and orientation of each bone model in a 3D space. Collisions among the reconstructed bones were also evaluated to avoid penetration. Using the biplanar X-ray fluoroscopic images and the proposed reconstruction methodology based on CT, we reconstructed 3D movements of the calcaneus, talus, navicular and cuboid when a human cadaveric foot walked on a flat surface using the simulator. The surface models of the four bones were successfully matched with the corresponding fluoroscopic images and the joint movements were quantified and visualized. The present methodology must be undergone further evaluation, but the proposed framework may serve as an effective tool for understanding the morphofunctional roles of the human foot structure during walking.

In this paper, a wireless human motion monitoring system based on joint angle sensors and smart shoes is introduced. An inertial measurement unit (IMU) is employed in a joint angle sensor to estimate the lower-extremity joint rotation in three dimensions. Four pressure sensors are embedded in a smart shoe to measure the distribution of ground contact forces (GCFs). Zig-bee and Bluetooth modules are combined with the joint angle sensors and smart shoes respectively to make the whole system wireless. It is shown that gait phase and step length can be calculated based on the raw sensor data for gait analysis. To provide visual feedback to the users, with the consent of Apple Inc., an user interface application is developed on an iPad. Experimental results are obtained from both a healthy subject and a stroke patient for comparison. Some discussions are made about the potential use of this system in a clinical environment.

Gait phase is an important state variable that helps exoskeletons provide assistance to users. Exoskeleton systems need to determine the assistance time and calculate the reference assistance force based on the gait phase. A large number of studies have used Finite State Machines (FSMs) or Adaptive Oscillators (AOs) to estimate gait phase. However, FSMs require manual calibration of gait events, and AOs can only estimate the phase of stable gait in real time. Both are difficult to adapt to irregular changes in gait. Deep learning algorithms can learn the nonlinear characteristics of kinematic data to estimate joint torque, gait phase, etc. The control algorithm based on Long Short-Term Memory (LSTM) Neural Network can automatically detect gait events according to the joint angle collected by Inertial Measurement Unit (IMU), effectively responding to dynamic changes in gait. This paper designed a Light-Weight Long Short-Term Memory (LWLSTM) Neural Network to estimate gait phase precisely and reduce training time of LSTM model. The experimental results showed that the Root Mean Square Error (RMSE) of LWLSTM was 8. 0 0 3 3%, quite lower than Convolutional Neural Network (CNN) and AO.

Walking is a fundamental human activity, and analyzing its complexities is essential for understanding gait abnormalities and musculoskeletal disorders. This article delves into the classification of gait phases using advanced machine learning techniques, specifically focusing on dividing these phases into five distinct subphases. The study utilizes data from 100 individuals obtained from an open-access platform and employs two distinct training methodologies. The first approach adopts stratified random sampling, where 80% of the data from each subphase are allocated for training and 20% for testing. The second approach involves participant-based splitting, training on data from 80% of the individuals and testing on the remaining 20%. Preprocessing methods such as Min–Max Scaling (MMS), Standard Scaling (SS), and Principal Component Analysis (PCA) were applied to the dataset to ensure optimal performance of the machine learning models. Several algorithms were implemented, including k-Nearest Neighbors (k-NNs), Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), Naive Bayes (Gaussian, Bernoulli, and Multinomial) (NB), Linear Discriminant Analysis (LDA), and Quadratic Discriminant Analysis (QDA). The models were rigorously evaluated using performance metrics like cross-validation score, Mean Squared Error (MSE), Root Mean Squared Error (RMSE), accuracy, and R2 score, offering a comprehensive assessment of their effectiveness in classifying gait phases. In the five subphases analysis, RF again performed strongly with a 94.95% accuracy, an RMSE of 0.4461, and an R2 score of 90.09%, demonstrating robust performance across all scaling methods.

The estimation of human gait phase has recently been used to parameterize continuous lower limb exoskeleton controllers. By deriving the gait phase estimation directly from the user's biological signals (i.e., joint/segment angles), the user can be granted an avenue to control their progression through the gait cycle. The inherent challenge for this strategy is the high accuracy required, as the estimated gait phase determines the timing of the device's assistive and resistive torques. Poor device timing can result in desynchronization of the device with the user. To address this challenge, this paper presents a novel method for using the information from the contralateral leg to improve the estimation accuracy. Simulation results showed a 75.2% reduction in estimation error relative to the state of the art method for able-bodied individuals. Additionally, an average of 47.6% reduction in estimation error was observed among data from three hemiparetic stroke patients. Across both able bodied and hemiparetic subjects, an estimation RMSE of 2.28% was shown. The evaluation of this technique displays its promise for the control of rehabilitative robotics.Clinical Relevance-This improvement to lower limb exoskeleton control aims to encourage greater user participation in robot assisted gait training for stroke rehabilitation.

Sensors and measurement systems is an introduction to microsensors for engineering students in the final undergraduate or early graduate level, technicians who wants to know more about the systems they are using, and anybody curious enough to know what microsystems and microsensors can do. The book discusses five families of sensors: - Thermal sensors - Force and pressure sensors- Inertial sensors - Magnetic field sensors- Flow sensorsFor each sensor, theoretical, technology and application aspects are examined. The sensor function is modelled to understand sensitivity, resolution and noise. We ask ourselves: What do we want to measure? What are possible applications? How are the sensor chips made in the cleanroom? How are they mounted and integrated in a system?After reading this book, you should be able to:- Understand important thermal, mechanical, inertial and magnetic sensors- Work with characterization parameters for sensors- Choose sensors for a given application and apply them- Understand micromachining technologies for sensors

Sensor-based human action recognition (HAR) is considered to have broad practical prospects. It applies to wearable devices to collect plantar pressure or acceleration information at human joints during human actions, thereby identifying human motion patterns. Existing related works have mainly focused on improving recognition accuracy, and have rarely considered energy-efficient management of portable HAR systems. Considering the high sensitivity and energy harvesting ability of triboelectric nanogenerators (TENGs), in this research a TENG which achieved output performance of 9.98 mW/cm2 was fabricated using polydimethylsiloxane and carbon nanotube film for sensor-based HAR as a wearable sensor. Considering real-time identification, data are acquired using a sliding window approach. However, the classification accuracy is challenged by quasi-periodic characteristics of the intercepted sequence. To solve this problem, compensatory dynamic time warping (C-DTW) is proposed, which adjusts the DTW result based on the proportion of points separated by small distances under DTW alignment. Our simulation results show that the classification accuracy of C-DTW is higher than that of DTW and its improved versions (e.g., WDTW, DDTW and softDTW), with almost the same complexity. Moreover, C-DTW is much faster than shapeDTW under the same classification accuracy. Without loss of generality, the performance of the existing DTW versions can be enhanced using the compensatory mechanism of C-DTW.

Monitoring gait quality in daily activities through wearable sensors has the potential to improve medical assessment in Parkinson’s Disease (PD). In this study, four gait partitioning methods, two based on thresholds and two based on a machine learning approach, considering the four-phase model, were compared. The methods were tested on 26 PD patients, both in OFF and ON levodopa conditions, and 11 healthy subjects, during walking tasks. All subjects were equipped with inertial sensors placed on feet. Force resistive sensors were used to assess reference time sequence of gait phases. Goodness Index (G) was evaluated to assess accuracy in gait phases estimation. A novel synthetic index called Gait Phase Quality Index (GPQI) was proposed for gait quality assessment. Results revealed optimum performance (G < 0.25) for three tested methods and good performance (0.25 < G < 0.70) for one threshold method. The GPQI resulted significantly higher in PD patients than in healthy subjects, showing a moderate correlation with clinical scales score. Furthermore, in patients with severe gait impairment, GPQI was found higher in OFF than in ON state. Our results unveil the possibility of monitoring gait quality in PD through real-time gait partitioning based on wearable sensors.

The absence of a midtarsal break has long been regarded as a derived feature of the human foot. Humans possess a rigid midfoot that acts as an efficient lever during the propulsive phase of bipedal gait. Non-human primates, in contrast, have a more mobile midfoot that is adaptive for tree climbing. Here, we report plantar pressure and video evidence that a small percentage of modern humans (n = 32/398) possess both elevated lateral midfoot pressures and even exhibit midfoot dorsiflexion characteristic of a midtarsal break. Those humans with a midtarsal break had on average a significantly flatter foot than those without. Midtarsal breakers also had significantly more medial weight transfer (pronation) during the stance phase of gait than those without this midfoot mobility. These data are in accordance with Elftman (Clin Orthop 16 (1960) 41-45) who suggested that pronation aligns the axes of the transverse tarsal joint, permitting elevated midfoot mobility.

Pose estimation of human motor skills such as bicycling in natural environments is challenging because of highly-dimensional human motion. In this paper, we present a dynamic rider/bicycle pose estimation scheme that can be used in outdoor environments. The proposed estimation scheme is based on the integration of the rider/bicycle dynamic model with the measurements from the force sensor and inertial measurement units (IMU). We take advantages of the attractive properties of both the force and IMU sensors in the design, that is, the force measurements do not suffer drifting while the IMU measurements generate real-time attitude and acceleration information. The rider/bicycle dynamic model provides an underlying relationship between the force and the IMU measurements. We demonstrate the effectiveness and performance of the pose estimation design through extensive bicycle riding experiments.