Fast Wearable Sensor–Based Foot–Ground Contact Phase Classification Using a Convolutional Neural Network with Sliding-Window Label Overlapping

Jeon, Haneul; Kim, Sang Lae; Kim, Soyeon; Lee, Donghun

doi:10.3390/s20174996

Cited by 13 publications

(12 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In 2020 another approach using a CNN was presented by Jeon et al [ 16 ] reaching 99.8% accuracy, but having a huge drop in accuracy when testing with unseen data. In this paper, a new ConvLSTM approach is presented to automatically divide the time series into its gait phases, with higher granularity and individuality compared to other papers before.…”

Section: Introductionmentioning

confidence: 99%

Deep Convolutional and LSTM Networks on Multi-Channel Time Series Data for Gait Phase Recognition

Kreuzer

Münz

2021

Sensors

View full text Add to dashboard Cite

With an ageing society comes the increased prevalence of gait disorders. The restriction of mobility leads to a considerable reduction in the quality of life, because associated falls increase morbidity and mortality. Consideration of gait analysis data often alters surgical recommendations. For that reason, the early and systematic diagnostic treatment of gait disorders can spare a lot of suffering. As modern gait analysis systems are, in most cases, still very costly, many patients are not privileged enough to have access to comparable therapies. Low-cost systems such as inertial measurement units (IMUs) still pose major challenges, but offer possibilities for automatic real-time motion analysis. In this paper, we present a new approach to reliably detect human gait phases, using IMUs and machine learning methods. This approach should form the foundation of a new medical device to be used for gait analysis. A model is presented combining deep 2D-convolutional and LSTM networks to perform a classification task; it predicts the current gait phase with an accuracy of over 92% on an unseen subject, differentiating between five different phases. In the course of the paper, different approaches to optimize the performance of the model are presented and evaluated.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep Convolutional and LSTM Networks on Multi-Channel Time Series Data for Gait Phase Recognition

Kreuzer

Münz

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…In their works, ref. [17] used windowed data as well to perform training on a two-dimensional format matrix. Related works such as [18,19] showed that the window size can be chosen based on a compromise between the accuracy and the training time, and for smaller values of the window size, we get a higher accuracy.…”

Section: Data Preparationmentioning

confidence: 99%

“…Models such as Naive Bayes were also used in the context of the study [30]. Hence, deep neural architectures can be used whether for the same configuration or for pre-processed and raw signals, especially CNN models that are commonly used and have shown their efficiency on raw data through the works of [17]. Furthermore, CNN architectures can also be used on sequential inputs by applying one-dimensional convolutional layers to each component of the different signals; ref.…”

Section: Neural Architecturesmentioning

confidence: 99%

Foot-to-Ground Phases Detection: A Comparison of Data Representation Formatting Methods with Respect to Adaption of Deep Learning Architectures

et al. 2022

View full text Add to dashboard Cite

Identifying the foot stance and foot swing phases, also known as foot-to-ground (FTG) detection, is a branch of Human Activity Recognition (HAR). Our study aims to detect two main phases of the gait (i.e., foot-off and foot-contact) corresponding to the moments when each foot is in contact with the ground or not. This will allow the medical professionals to characterize and identify the different phases of the human gait and their respective patterns. This detection process is paramount for extracting gait features (e.g., step width, stride width, gait speed, cadence, etc.) used by medical experts to highlight gait anomalies, stance issues, or any other walking irregularities. It will be used to assist health practitioners with patient monitoring, in addition to developing a full pipeline for FTG detection that would help compute gait indicators. In this paper, a comparison of different training configurations, including model architectures, data formatting, and pre-processing, was conducted to select the parameters leading to the highest detection accuracy. This binary classification provides a label for each timestamp informing whether the foot is in contact with the ground or not. Models such as CNN, LSTM, and ConvLSTM were the best fits for this study. Yet, we did not exclude DNNs and Machine Learning models, such as Random Forest and XGBoost from our work in order to have a wide range of possible comparisons. As a result of our experiments, which included 27 senior participants who had a stroke in the past wearing IMU sensors on their ankles, the ConvLSTM model achieved a high accuracy of 97.01% for raw windowed data with a size of 3 frames per window, and each window was formatted to have two superimposed channels (accelerometer and gyroscope channels). The model was trained to have the best detection without any knowledge of the participants’ personal information including age, gender, health condition, the type of activity, or the used foot. In other words, the model’s input data only originated from IMU sensors. Overall, in terms of FTG detection, the combination of the ConvLSTM model and the data representation had an important impact in outperforming other start-of-the-art configurations; in addition, the compromise between the model’s complexity and its accuracy is a major asset for deploying this model and developing real-time solutions.

show abstract

“…Since the lower limb exoskeleton is powered by motor-driven joints, including the hip and knee f/e, the planned Cartesian position coordinates of the ankle and hip joints should be converted to the joint space through inverse kinematics as shown in Equations ( 10)- (14). Consecutive steps of walking are simulated with a sequence of stride lengths, L stride = [1245, 1150, 1050, 950, 850, 750, 650].…”

Section: Gait Planning Algorithm Simulationmentioning

confidence: 99%

“…Therefore, gait planning online or offline with additional measured information has been studied in recent years. Jeon et al presented a fast wearable sensor-based gait phase classification method with the help of a convolutional neural network, which represents human-machine walking intention and is useful for exoskeleton motion control [ 14 ]. The polymer optical fiber sensors reported by Leal-Junior et al show promising application scenarios for soft and wearable gait measurement for exoskeleton online gait planning [ 15 ].…”

Section: Introductionmentioning

confidence: 99%

Lower Limb Exoskeleton Gait Planning Based on Crutch and Human-Machine Foot Combined Center of Pressure

Yang

Zhang

et al. 2020

Sensors

View full text Add to dashboard Cite

With the help of wearable robotics, the lower limb exoskeleton becomes a promising solution for spinal cord injury (SCI) patients to recover lower body locomotion ability. However, fewer exoskeleton gait planning methods can meet the needs of patient in real time, e.g., stride length or step width, etc., which may lead to human-machine incoordination, limit comfort, and increase the risk of falling. This work presents a human-exoskeleton-crutch system with the center of pressure (CoP)-based gait planning method to enable the balance control during the exoskeleton-assisted walking with crutches. The CoP generated by crutches and human-machine feet makes it possible to obtain the overall stability conditions of the system in the process of exoskeleton-assisted quasi-static walking, and therefore, to determine the next stride length and ensure the balance of the next step. Thus, the exoskeleton gait is planned with the guidance of stride length. It is worth emphasizing that the nominal reference gait is adopted as a reference to ensure that the trajectory of the swing ankle mimics the reference one well. This gait planning method enables the patient to adaptively interact with the exoskeleton gait. The online gait planning walking tests with five healthy volunteers proved the method’s feasibility. Experimental results indicate that the algorithm can deal with the sensed signals and plan the landing point of the swing leg to ensure balanced and smooth walking. The results suggest that the method is an effective means to improve human–machine interaction. Additionally, it is meaningful for the further training of independent walking stability control in exoskeletons for SCI patients with less assistance of crutches.

show abstract

Fast Wearable Sensor–Based Foot–Ground Contact Phase Classification Using a Convolutional Neural Network with Sliding-Window Label Overlapping

Cited by 13 publications

References 30 publications

Deep Convolutional and LSTM Networks on Multi-Channel Time Series Data for Gait Phase Recognition

Deep Convolutional and LSTM Networks on Multi-Channel Time Series Data for Gait Phase Recognition

Foot-to-Ground Phases Detection: A Comparison of Data Representation Formatting Methods with Respect to Adaption of Deep Learning Architectures

Lower Limb Exoskeleton Gait Planning Based on Crutch and Human-Machine Foot Combined Center of Pressure

Contact Info

Product

Resources

About