Gait Events Research Articles

Evaluating the Impact of IMU Sensor Location and Walking Task on Accuracy of Gait Event Detection Algorithms.

There are several algorithms that use the 3D acceleration and/or rotational velocity vectors from IMU sensors to identify gait events (i.e., toe-off and heel-strike). However, a clear understanding of how sensor location and the type of walking task effect the accuracy of gait event detection algorithms is lacking. To address this knowledge gap, seven participants were recruited (4M/3F; 26.0 ± 4.0 y/o) to complete a straight walking task and obstacle navigation task while data were collected from IMUs placed on the foot and shin. Five different commonly used algorithms to identify the toe-off and heel-strike gait events were applied to each sensor location on a given participant. Gait metrics were calculated for each sensor/algorithm combination using IMUs and a reference pressure sensing walkway. Results show algorithms using medial-lateral rotational velocity and anterior-posterior acceleration are fairly robust against different sensor locations and walking tasks. Certain algorithms applied to heel and lower lateral shank sensor locations will result in degraded algorithm performance when calculating gait metrics for curved walking compared to straight overground walking. Understanding how certain types of algorithms perform for given sensor locations and tasks can inform robust clinical protocol development using wearable technology to characterize gait in both laboratory and real-world settings.

FSR and IMU sensors-based human gait phase detection and its correlation with EMG signal for different terrain walk

PurposeThe purpose of this paper is to present gait analysis for five different terrains: level ground, ramp ascent, ramp descent, stair ascent and stair descent.Design/methodology/approachGait analysis has been carried out using a combination of the following sensors: force-sensitive resistor (FSR) sensors fabricated in foot insole to sense foot pressure, a gyroscopic sensor to detect the angular velocity of the shank and MyoWare electromyographic muscle sensors to detect muscle’s activities. All these sensors were integrated around the Arduino nano controller board for signal acquisition and conditioning purposes. In the present scheme, the muscle activities were obtained from the tibialis anterior and medial gastrocnemius muscles using electromyography (EMG) electrodes, and the acquired EMG signals were correlated with the simultaneously attained signals from the FSR and gyroscope sensors. The nRF24L01+ transceivers were used to transfer the acquired data wirelessly to the computer for further analysis. For the acquisition of sensor data, a Python-based graphical user interface has been designed to analyze and display the processed data. In the present paper, the authors got motivated to design and develop a reliable real-time gait phase detection technique that can be used later in designing a control scheme for the powered ankle-foot prosthesis.FindingsThe effectiveness of the gait phase detection was obtained in an open environment. Both off-line and real-time gait events and gait phase detections were accomplished for the FSR and gyroscopic sensors. Both sensors showed their usefulness for detecting the gait events in real-time, i.e. within 10 ms. The heuristic rules and a zero-crossing based-algorithm for the shank angular rate correctly identified all the gait events for the locomotion in all five terrains.Practical implicationsThis study leads to an understanding of human gait analysis for different types of terrains. A real-time standalone system has been designed and realized, which may find application in the design and development of ankle-foot prosthesis having real-time control feature for the above five terrains.Originality/valueThe noise-free data from three sensors were collected in the same time frame from both legs using a wireless sensor network between two transmitters and a single receiver. Unlike the data collection using a treadmill in a laboratory environment, this setup is useful for gait analysis in an open environment for different terrains.

Wearable Knee Assistive Devices for Kneeling Tasks in Construction

Construction workers regularly perform tasks that require kneeling, crawling, and squatting. Working in awkward kneeling postures for prolonged time periods can lead to knee pain, injuries, and osteoarthritis. In this article, we present lightweight, wearable sensing, and knee assistive devices for construction workers during kneeling and squatting tasks. Analysis of kneeling on level and sloped surfaces (0 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> , 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> , and 20 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">°</sup> ) is performed for single- and double-leg kneeling tasks. Measurements from the integrated inertial measurement units are used for real-time gait detection and lower limb pose estimation. Detected gait events and pose estimation are used to control the assistive knee joint torque provided by lightweight exoskeletons with powerful quasi-direct drive actuation. Human subject experiments are conducted to validate the effectiveness of the proposed analysis and control design. The results show reduction in knee extension/flexion muscle activation (up to 39%) during stand-to-kneel and kneel-to-stand tasks. Knee-ground contact forces/pressures are also reduced (up to 15%) under robotic assistance during single-leg kneeling. Increasing assistive knee torque shows redistribution of the subject's weight from the knee in contact with the ground to both supporting feet. The proposed system provides an enabling intervention to potentially reduce musculoskeletal injury risks of construction workers.

Joint angle estimation with wavelet neural networks

This paper presents a wavelet neural network (WNN) based method to reduce reliance on wearable kinematic sensors in gait analysis. Wearable kinematic sensors hinder real-time outdoor gait monitoring applications due to drawbacks caused by multiple sensor placements and sensor offset errors. The proposed WNN method uses vertical Ground Reaction Forces (vGRFs) measured from foot kinetic sensors as inputs to estimate ankle, knee, and hip joint angles. Salient vGRF inputs are extracted from primary gait event intervals. These selected gait inputs facilitate future integration with smart insoles for real-time outdoor gait studies. The proposed concept potentially reduces the number of body-mounted kinematics sensors used in gait analysis applications, hence leading to a simplified sensor placement and control circuitry without deteriorating the overall performance.

Spatiotemporal Analysis by Deep Learning of Gait Signatures From Floor Sensors

The recognition of gait pattern variation is of high importance to various industrial and commercial applications, including security, sport, virtual reality, gaming, robotics, medical rehabilitation, mental illness diagnosis, space exploration, and others. The purpose of this paper is to study the nature of gait variability in more detail, by identifying gait intervals responsible for gait pattern variations in individuals, as well as between individuals, using cognitive demanding tasks. This work uses deep learning methods for sensor fusion of 116 plastic optical fiber (POF) distributed sensors for gait recognition. The floor sensor system captures spatiotemporal samples due to varying ground reaction force (GRF) in multiples of up to 4 uninterrupted steps on a continuous 2×1 m area. We demonstrate classifications of gait signatures, achieving up to 100% F1-score with Convolutional Neural Networks (CNN), in the context of gait recognition of 21 subjects, with imposters and clients. Classifications under cognitive load, induced by 4 different dual tasks, manifested lower F1-scores. Layer-Wise Relevance Propagation (LRP) methods are employed to decompose a trained neural network prediction to relevant standard events in the gait cycle, by generating a “heat map” over the input used for classification. This allows valuable insight into which parts of the gait spatiotemporal signal have the heaviest influence on the gait classification and consequently, which gait events, such as heel strike or toe-off, are mostly affected by cognitive load.

26 Detecting equine gait events in field conditions

As interest in analyzing equine gait in the real world grows, so does need for quantitative methods validated under field conditions. Gait analysis is predicated on accurate and precise detection of gait events– hoof on and off. Several methods of detection have been proposed which use inertial measurement units (IMUs); the most accurate is based on hoof mounted IMUs but attachment to the hooves is not always convenient. The aim of this research was thus to propose and test a novel gait event detection method using pastern mounted IMUs. Previous studies only focused on one hard surface; asphalt, grass and sand were investigated here. Eight geldings and 3 mares (mean(SD) height 154(21)cm and age 12(8) yr) including Irish Sport Horses, Warmblood and native breeds wore IMUs (Shimmer, 200Hz) to be led in a 25m straight line at walk and trot on asphalt, grass and sand (3 passes per gait). Stance durations (T) were calculated from hoof on and off obtained using both a reference (hoof) and novel (pastern) method. Error was assessed in terms of mean difference between pastern (Tp) and hoof (Th) calculated stance durations (accuracy) and SD of these (precision), as a percentage of total stride duration. Repeated measures ANOVA was used to compare errors under different conditions. Accuracy (Table 1) remained below 1% of a stride duration for most cases. Performance of the method was consistent on all surfaces, with no significant differences in most cases ( P = 0.4 fore and P = 0.08 hindlimb walk; P = 0.3 hindlimb trot). For forelimbs at trot differences between grass and sand compared with asphalt were significant ( P = 0.004 and P < 0.0001) but the magnitude of these, 0.5 and 0%, were negligible. To conclude, the novel method effectively detected gait events on 3 common surfaces for the fore and hindlimbs at walk and trot across multiple breeds, with high accuracy and precision on all surfaces. This will benefit future studies under field conditions, allowing reliable event detection on different surfaces.

Trabecular bone properties in the ilium of the Middle Paleolithic/Middle Stone Age Border Cave 3 Homo sapiens infant and the onset of independent gait

The Border Cave 3 (BC3) infant skeleton has been understudied, despite its importance as an example of a well-preserved and fairly complete immature skeleton of early Homo sapiens which potentially provides a rare window into various aspects of ontogenetic development, including locomotor activity (e.g., timing of gait events). Trabecular structure in the BC3 ilium was evaluated to investigate whether it matches that of an equivalently aged infant from a postindustrialized society. Microcomputed tomography (μCT) scans were acquired from the BC3 infant and from an ontogenetic series of 25 postindustrial infants that were divided into three age classes (ACs) ranging from neonates to toddlers (<36 months). All ilia were qualitatively compared and then digitally subdivided into 10 volumes of interest (VOIs) based on anatomical reference points. The VOIs were quantified and ontogenetic differences in trabecular structure were statistically evaluated. Across the comparative ontogenetic series, trabecular architectural properties overlapped in all regions. However, trabecular thickness increased significantly after the first year of life. The BC3 infant demonstrated generally similar trabecular structure to that observed in the age-equivalent postindustrial infants (AC2), including relatively strong development of the trabecular chiasma qualitatively. However, some interesting distinctions were observed in BC3, such as low strut thickness compared with infants from the postindustrial sample, that bear further exploration in future studies. Evaluation of only one individual from the Middle Stone Age (MSA), coupled with the relatively small comparative sample, limit our ability to distinguish more meaningful biological differences in trabecular structure throughout ontogeny from idiosyncratic characteristics. Nonetheless, results of this study extend ongoing research on infant locomotor and morphological development to archeological populations in the Middle Stone Age. Further cross-cultural studies consisting of larger comparative postindustrial samples may provide additional information on trabecular structure in the infant ilium during this important developmental timeframe.

Locomotion Mode Recognition for Walking on Three Terrains Based on sEMG of Lower Limb and Back Muscles

Gait phase detection on different terrains is an essential procedure for amputees with a lower limb assistive device to restore walking ability. In the present study, the intent recognition of gait events on three terrains based on sEMG was presented. The class separability and robustness of time, frequency, and time-frequency domain features of sEMG signals from five leg and back muscles were quantitatively evaluated by statistical analysis to select the best features set. Then, ensemble learning method that combines the outputs of multiple classifiers into a single fusion-produced output was implemented. The results obtained from data collected from four human participants revealed that the light gradient boosting machine (LightGBM) algorithm has an average accuracy of 93.1%, a macro-F1 score of 0.929, and a calculation time of prediction of 15 ms in discriminating 12 different gait phases on three terrains. This was better than traditional voting-based multiple classifier fusion methods. LightGBM is a perfect choice for gait phase detection on different terrains in daily life.

Measuring Gait Velocity and Stride Length with an Ultrawide Bandwidth Local Positioning System and an Inertial Measurement Unit.

One possible modality to profile gait speed and stride length includes using wearable technologies. Wearable technology using global positioning system (GPS) receivers may not be a feasible means to measure gait speed. An alternative may include a local positioning system (LPS). Considering that LPS wearables are not good at determining gait events such as heel strikes, applying sensor fusion with an inertial measurement unit (IMU) may be beneficial. Speed and stride length determined from an ultrawide bandwidth LPS equipped with an IMU were compared to video motion capture (i.e., the “gold standard”) as the criterion standard. Ninety participants performed trials at three self-selected walk, run and sprint speeds. After processing location, speed and acceleration data from the measurement systems, speed between the last five meters and stride length in the last stride of the trial were analyzed. Small biases and strong positive intraclass correlations (0.9–1.0) between the LPS and “the gold standard” were found. The significance of the study is that the LPS can be a valid method to determine speed and stride length. Variability of speed and stride length can be reduced when exploring data processing methods that can better extract speed and stride length measurements.

Adaptive Pose Estimation for Gait Event Detection Using Context-Aware Model and Hierarchical Optimization

To understand daily events accurately, adaptive pose estimation (APE) systems require a robust context-aware model and optimal feature selection methods. In this paper, we propose a novel gait event detection (GED) system that consists of saliency silhouette detection, a robust body parts model and a 2D stick-model followed by a hierarchical optimization algorithm. Furthermore, the most prominent context-aware features such as energy, 0–180° intensity and distinct moveable features are proposed by focusing on invariant and localized characteristics of human postures in different event classes. Finally, we apply Grey Wolf optimization and a genetic algorithm to discriminate complex postures and to provide appropriate labels to each event. In order to evaluate the performance of proposed GED, two public benchmark datasets, UCF101 and YouTube, are examined via the n-fold cross validation method. For the two benchmark datasets, our proposed method detects the human body key points with 82.4% and 83.2% accuracy respectively. Also, it extracts the context-aware features and finally recognizes the gait events with 82.6% and 85.0% accuracy, respectively. Compared with other well-known statistical and state-of-the-art methods, our proposed method outperforms other similarly tasked methods in terms of posture detection and recognition accuracy.

Assessment of spatiotemporal gait parameters using a deep learning algorithm-based markerless motion capture system

Spatiotemporal parameters can characterize the gait patterns of individuals, allowing assessment of their health status and detection of clinically meaningful changes in their gait. Video-based markerless motion capture is a user-friendly, inexpensive, and widely applicable technology that could reduce the barriers to measuring spatiotemporal gait parameters in clinical and more diverse settings. Two studies were performed to determine whether gait parameters measured using markerless motion capture demonstrate concurrent validity with those measured using marker-based motion capture and a pressure-sensitive gait mat. For the first study, thirty healthy young adults performed treadmill gait at self-selected speeds while marker-based motion capture and synchronized video data were recorded simultaneously. For the second study, twenty-five healthy young adults performed over-ground gait at self-selected speeds while footfalls were recorded using a gait mat and synchronized video data were recorded simultaneously. Kinematic heel-strike and toe-off gait events were used to identify the same gait cycles between systems. Nine spatiotemporal gait parameters were measured by each system and directly compared between systems. Measurements were compared using Bland-Altman methods, mean differences, Pearson correlation coefficients, and intraclass correlation coefficients. The results indicate that markerless measurements of spatiotemporal gait parameters have good to excellent agreement with marker-based motion capture and gait mat systems, except for stance time and double limb support time relative to both systems and stride width relative to the gait mat. These findings indicate that markerless motion capture can adequately measure spatiotemporal gait parameters of healthy young adults during treadmill and over-ground gait.

The Smart-Insole Dataset: Gait Analysis Using Wearable Sensors with a Focus on Elderly and Parkinson's Patients.

Gait analysis is crucial for the detection and management of various neurological and musculoskeletal disorders. The identification of gait events is valuable for enhancing gait analysis, developing accurate monitoring systems, and evaluating treatments for pathological gait. The aim of this work is to introduce the Smart-Insole Dataset to be used for the development and evaluation of computational methods focusing on gait analysis. Towards this objective, temporal and spatial characteristics of gait have been estimated as the first insight of pathology. The Smart-Insole dataset includes data derived from pressure sensor insoles, while 29 participants (healthy adults, elderly, Parkinson’s disease patients) performed two different sets of tests: The Walk Straight and Turn test, and a modified version of the Timed Up and Go test. A neurologist specialized in movement disorders evaluated the performance of the participants by rating four items of the MDS-Unified Parkinson’s Disease Rating Scale. The annotation of the dataset was performed by a team of experienced computer scientists, manually and using a gait event detection algorithm. The results evidence the discrimination between the different groups, and the verification of established assumptions regarding gait characteristics of the elderly and patients suffering from Parkinson’s disease.

Unrestricted stride detection during stair climbing using IMUs

Stride detection, or the identification of the initial (IC) and terminal contact (TC) of the feet while walking, is important for gait analysis. Automatic stride detection based only on kinematic data is challenging, even when using portable, low-cost, user-friendly Inertial Measurement Units (IMUs). Although there are algorithms for straight walking available, they are often not applicable to other movement patterns. Furthermore, these algorithms are based on the use of different IMUs placed on different locations of the body with different pre-processing filters and rely on analyzing different measurement signals. Therefore, it is difficult to apply existing algorithms for specific study settings.To achieve a new algorithm, thirty-five healthy participants were analyzed during walking and stair climbing while kinematic motion data was measured using the IMU system MyoMotion. Based on the analysis of different published methods for IC and TC detection, a new robust stride detection algorithm was developed and validated in comparison with two different algorithms. From this, it was determined that the newly developed algorithm was successful in automatic stride detection during walking and ascending/ descending stairs with 100% detected gait events, while the other algorithms failed during stair climbing with only 44% and 91% detected gait events.

Wearable Sensor-Based Real-Time Gait Detection: A Systematic Review.

Gait analysis has traditionally been carried out in a laboratory environment using expensive equipment, but, recently, reliable, affordable, and wearable sensors have enabled integration into clinical applications as well as use during activities of daily living. Real-time gait analysis is key to the development of gait rehabilitation techniques and assistive devices such as neuroprostheses. This article presents a systematic review of wearable sensors and techniques used in real-time gait analysis, and their application to pathological gait. From four major scientific databases, we identified 1262 articles of which 113 were analyzed in full-text. We found that heel strike and toe off are the most sought-after gait events. Inertial measurement units (IMU) are the most widely used wearable sensors and the shank and foot are the preferred placements. Insole pressure sensors are the most common sensors for ground-truth validation for IMU-based gait detection. Rule-based techniques relying on threshold or peak detection are the most widely used gait detection method. The heterogeneity of evaluation criteria prevented quantitative performance comparison of all methods. Although most studies predicted that the proposed methods would work on pathological gait, less than one third were validated on such data. Clinical applications of gait detection algorithms were considered, and we recommend a combination of IMU and rule-based methods as an optimal solution.

A Novel Method for Stride Length Estimation Using Wireless Foot Sensor Module

This work presents a novel method for the estimation of stride length (SL) from inertial sensor-based wireless foot sensor module (WFSM). SL variability is a widely used and clinically relevant spatial gait parameter. This work proposes and validates a novel method for computing SL by measuring the foot inclination angle and single-axis acceleration measured from WFSM. The foot inclination angle is used for detecting gait events as well as for gravity compensation to estimate the real acceleration of the foot. This approach thus requires minimum input information, i.e. foot angle trajectory and acceleration in only sagittal plane. The results are validated on a cohort of healthy subjects against a standard gait analysis platform. A set of trial data on 10 healthy volunteers were also analysed for outdoor walking to evaluate the performance against previously reported works. The proposed approach has the potential for applications involving real-time gait measurements.

Real-Time Gait Phase Estimation for Robotic Hip Exoskeleton Control During Multimodal Locomotion.

We developed and validated a gait phase estimator for real-time control of a robotic hip exoskeleton during multimodal locomotion. Gait phase describes the fraction of time passed since the previous gait event, such as heel strike, and is a promising framework for appropriately applying exoskeleton assistance during cyclic tasks. A conventional method utilizes a mechanical sensor to detect a gait event and uses the time since the last gait event to linearly interpolate the current gait phase. While this approach may work well for constant treadmill walking, it shows poor performance when translated to overground situations where the user may change walking speed and locomotion modes dynamically. To tackle these challenges, we utilized a convolutional neural network-based gait phase estimator that can adapt to different locomotion mode settings to modulate the exoskeleton assistance. Our resulting model accurately predicted the gait phase during multimodal locomotion without any additional information about the user's locomotion mode, with a gait phase estimation RMSE of 5.04 ± 0.79%, significantly outperforming the literature standard (p < 0.05). Our study highlights the promise of translating exoskeleton technology to more realistic settings where the user can naturally and seamlessly navigate through different terrain settings.

Control and Evaluation of Body Weight Support Walker for Overground Gait Training

Gait rehabilitation training under robot-assisted partial Body Weight Support (BWS) is a promising technique that helps patients who suffer from a traumatic or congenital brain injury like stroke or cerebral palsy to become independent in daily life activities. In recent years, robotic BWS systems have been widely studied, where the BWS is provided by the robot while the user walks on devices fixed to the environment such as treadmills, thereby gaining functional benefits such as improved gait symmetry and increased walking speed. On the other hand, mobile BWS robots that allow conventional overground walking with well-designed control strategies have been less researched, limiting the widespread adoption of robotic rehabilitation because of the cost effectiveness of fixed robotic devices and poor portability. To address this problem, in our previous study, we developed a mobile BWS walker that allows for overground walking under variable levels of BWS. In this letter, we introduce a system architecture that integrates the walker with a pair of instrumented shoes and discuss different system control strategies including static and variable BWS control. When walking under Static-BWS (SBWS), the subject walks while a constant portion of his/her weight is supported, which might cause an unnatural gait. Using a Variable-BWS (VBWS), the provided BWS can be adjusted according to human's gait events to provide a more natural gait. This letter describes the system architecture, as well as experiments with able subjects that demonstrate the effectiveness of both the SBWS and VBWS control algorithms in relieving part of the subjects’ body weight. Specifically, by synchronizing the controlled BWS with human gait events, users were able to walk more naturally, particularly under a high level of BWS.

Detection of Equine Hoof Motion by Using a Hoof-Mounted Inertial Measurement Unit Sensor in Comparison to Examinations with an Optoelectronic Technique - A Pilot Study

This study aimed to determine whether a hoof-mounted inertial measurement unit (IMU) system would provide similar timing characteristics and step lengths, in walk and trot on different grounds, compared to those of an established optoelectronic motion system. The right hoof of one horse was equipped with an IMU sensor with an optical reference marker on top. Fifteen steps on firm ground, and five on penetrable ground, were performed at walk and trot. Level of agreement between the two measurement systems was determined by assessing the concordance correlation coefficients, accuracy, and precision of the duration of different gait events and step length. Timing characteristics and step length were in strong agreement between the two techniques for the majority of assessed parameters in walk and trot on both grounds. The agreement between techniques decreased for breakover and landing duration at trot on penetrable ground. Disparity between the measurement techniques was related to difficulties in accurately defining single parts of the stance phase with marker-based optoelectronic analysis on penetrable ground. Detailed examinations of different parts of the stance phase are more accurately performed using hoof-mounted IMU sensors. Results emphasise the great potential of IMU sensors for equine motion analysis in daily practice.

A Sensor-Based Multichannel FES System to Control Knee Joint and Reduce Stance Phase Asymmetry in Post-Stroke Gait.

Most of the studies using functional electrical stimulation (FES) in gait rehabilitation have been focused on correcting the drop foot syndrome. Using FES to control the knee joint in individuals with central nervous system (CNS) disorders could also play a key role in gait recovery: spasticity decrease, higher range of motion, positive effect on balance, limiting hyperextension and flexion in stance phase, reducing joint overload, etc. In stance phase, an accurate timing and a fine tuning of stimulation parameters are however required to provide a proper control of the knee stimulation while ensuring a safe and efficient support. In this study, 11 participants were equipped with inertial measurements units (IMU) and foot pressure insoles after supratentorial ischemic or hemorrhagic stroke, informing on knee angle and gait events used to online adapt FES during a 10 m walking protocol. Asymmetry of stance time and weight bearing were monitored as well as gait quality and physiological cost through a series of relevant markers. Vertical trunk motion has been significantly reduced during gait with FES (p-value = 0.038). Despite no significant improvement of stance phase asymmetry has been found, this preliminary work shows evidence of promising technical and rehabilitative potentials of a sensor-based multichannel FES system to control knee joint in post-stroke gait.

A novel fusion strategy for locomotion activity recognition based on multimodal signals

Detecting locomotion activities and gait phases are critical for recognizing movement intent and assisting humans in activities of daily living. This paper introduces a novel fusion strategy that analyzes multimodal signals (electromyograms, angular rates, and footswitches) to identify four lower limb locomotion activities (walking, running, stair ascent, and descent), four gait phases (push-up, stance, step-up, and swing) and detect gait events (heel strike and toe-off). This strategy uses different networks (convolutional neural network and Bi-directional Long Short-Term Memory) and a feature fusion method (Canonical Correlation Analysis) to perform the classification tasks and gait event detection, which transcends the limitations of conventional methods that only analyze a single signal. Then, an adaptive online algorithm is proposed, which helps the model to automatically fine-tune its parameters and gradually approach the convergence state. The offline experiment results on 20 healthy subjects show an average accuracy of 99.33 % and an average Matthews correlation coefficient of 99.56. Besides, with the help of the adaptive online algorithm, lower errors are shown in gait event detection, including 18.43±7.38 ms for heel strike and 22.25±7.06 ms for toe-off. Consequently, our proposed model can provide more accurate quantitative results than mainstream methods, giving it a competitive advantage for future applications.