Gait classification of knee osteoarthritis patients using shoe-embedded internal measurement units sensor.

BACKGROUND: Knee osteoarthritis negatively affects the gait of patients, especially that of elderly people. However, the assessment of wearable sensors in knee osteoarthritis patients has been under-researched. During clinical assessments, patients may change their gait patterns under the placebo effect, whereas wearable sensors can be used in any environment. METHODS: Sixty patients with knee osteoarthritis and 20 control subjects were included in the study. Wearing shoes with an IMU sensor embedded in the insoles, the participants were required to walk along a walkway. The sensor data were collected during the gait. To discriminate between healthy and knee osteoarthritis patients and to classify different subgroups of knee osteoarthritis patients (patients scheduled for surgery vs. patients not scheduled for surgery; bilateral knee osteoarthritis diagnosis vs. unilateral knee osteoarthritis diagnosis), we used a machine learning approach called the support vector machine. A total of 88 features were extracted and used for classification. FINDINGS: The patients vs. healthy participants were classified with 71% accuracy, 85% sensitivity, and 56% specificity. The "patients scheduled for surgery" vs. "patients not scheduled for surgery" were classified with 83% accuracy, 83% sensitivity, and 81% specificity. The bilateral knee osteoarthritis diagnosis vs. unilateral knee osteoarthritis diagnosis was classified with 81% accuracy, 75% sensitivity, and 79% specificity. INTERPRETATION: Gait analysis using wearable sensors and machine learning can discriminate between healthy and knee osteoarthritis patients and classify different subgroups with reasonable accuracy, sensitivity, and specificity. The proposed approach requires no complex gait factors and is not limited to controlled laboratory settings.

Differences in kinematic parameters during gait between the patients with knee osteoarthritis and healthy controls using an insole with a single inertial measurement unit: A case-control study.

BACKGROUND: An inertial measurement unit is small and lightweight, allowing patient measurements without physical constraints. This study aimed to determine the differences in kinematic parameters during gait using an insole with a single inertial measurement unit in healthy controls and on both sides in patients with knee osteoarthritis. METHODS: Twenty patients with knee osteoarthritis and 13 age-matched controls were included in this study. The participants walked at a self-selected speed and foot kinematics were measured during gait using an insole with a single inertial measurement unit. The right side of the healthy controls and both the affected and contralateral sides of patients with KOA were analyzed separately. FINDINGS: The foot extension angular velocity at toe-off was significantly reduced on the affected side than on the contralateral side (P < 0.001) and in healthy controls (P < 0.001). During the swing phase, foot posterior-anterior acceleration was significantly lower on the affected side than on the healthy controls (P = 0.005). Furthermore, despite a decrease in walking speed, foot superior-inferior acceleration at initial contact in patients was significantly lower on the contralateral side than in healthy controls (P = 0.0167), but not on the affected side (P = 0.344). INTERPRETATION: An insole with a single inertial measurement unit can detect differences in foot kinematics during gait between healthy controls and patients with knee osteoarthritis. Our findings indicate that patients with knee osteoarthritis exhibit dysfunction of push-off at toe-off and shock absorption at initial contact on the affected side.

Feature selection, construction and test of model for estimating lower extremity strength of older adults using foot motion measured by an in-shoe motion sensor.

Lower extremity strength (LES) is essential to support activities in daily living. To extend healthy life expectancy of elderly people, early detection of LES weakness is important. In this study, we challenge to develop a method for LES assessment in daily living via an in-shoe motion sensor (IMS). To construct the estimation model, we collected data from 62 subjects. We used the outcome of the five-times-sit-to-stand test to represent the performance of LES as the target variable. Predictors were constructed from the subjects' foot motions measured by the IMS during straight path walking. We used the leave-one-subject-out least absolute shrinkage and selection operator algorithm to select features and construct respective models for the males and females. As a result, the models achieved fair and a good intra-class correlation coefficient agreement between the true and estimation values, with mean absolute errors of 2.14 and 1.21 s (variation of 23.6 and 16.0%), respectively. To validate the models, we separately collected data from 45 subjects. The models successfully predicted 100% and 90% of the male and female subjects' data, respectively, which suggests the robustness of the constructed estimation models. The results suggested that LES can be identified more effectively in daily living by wearing an IMS, and the use of an IMS has the potential for future frailty and fall risk assessment applications.

Healthcare Application of In-Shoe Motion Sensor for Older Adults: Frailty Assessment Using Foot Motion during Gait.

Frailty poses a threat to the daily lives of healthy older adults, highlighting the urgent need for technologies that can monitor and prevent its progression. Our objective is to demonstrate a method for providing long-term daily frailty monitoring using an in-shoe motion sensor (IMS). We undertook two steps to achieve this goal. Firstly, we used our previously established SPM-LOSO-LASSO (SPM: statistical parametric mapping; LOSO: leave-one-subject-out; LASSO: least absolute shrinkage and selection operator) algorithm to construct a lightweight and interpretable hand grip strength (HGS) estimation model for an IMS. This algorithm automatically identified novel and significant gait predictors from foot motion data and selected optimal features to construct the model. We also tested the robustness and effectiveness of the model by recruiting other groups of subjects. Secondly, we designed an analog frailty risk score that combined the performance of the HGS and gait speed with the aid of the distribution of HGS and gait speed of the older Asian population. We then compared the effectiveness of our designed score with the clinical expert-rated score. We discovered new gait predictors for HGS estimation via IMSs and successfully constructed a model with an "excellent" intraclass correlation coefficient and high precision. Moreover, we tested the model on separately recruited subjects, which confirmed the robustness of our model for other older individuals. The designed frailty risk score also had a large effect size correlation with clinical expert-rated scores. In conclusion, IMS technology shows promise for long-term daily frailty monitoring, which can help prevent or manage frailty for older adults.

Identification of characteristics of foot position and angle during swing phase in fallers using principal component analysis.

Identifying the characteristics of fallers is important for preventing falls because such events may reduce quality of life. It has been reported that several variables related to foot positions and angles during gait (e.g., sagittal foot angle and minimum toe clearance) differ between fallers and non-fallers. However, examining such representative discrete variables may not be sufficient to detect crucial information, which may be contained in the large portions of unanalyzed data. Therefore, we aimed to identify the comprehensive characteristics of foot position and angle during the swing phase of gait in non-fallers and fallers using principal component analysis (PCA). Thirty non-fallers and 30 fallers were recruited for this study. We performed PCA to reduce the dimensions of foot positions and angles during the swing phase and obtained principal component scores (PCSs) for each principal component vector (PCV), which were then compared between groups. The results revealed that the PCS of PCV3 in fallers was significantly larger than that in non-fallers (p = 0.003, Cohen's d = 0.80). We reconstructed waveforms of foot positions and angles during the swing phase using PCV3 and our main findings can be summarized as follows. Compared to non-fallers, fallers have a 1) low average foot position in the z-axis (i.e., height) during the initial swing phase 2) small average foot angle in the x-axis (i.e., rotation in the sagittal plane), during the initial swing phase, and 3) large variability in foot position in the y-axis (i.e., anterior/posterior position) during the initial swing phase. We can conclude that these are characteristics of gait related to fallers. Therefore, our findings may be beneficial for evaluating fall risk during gait using a device such as a shoe- or insole-embedded inertial measurement unit.

Estimation of Hand Grip Strength Using Foot motion Measured by In-shoe Motion Sensor.

There is a strong need to assess frailty in daily living. Hand grip strength (HGS) has been proven to be a very important factor for identifying frailty, however it is always assessed under the guidance of facility clinicians. Our purpose is to demonstrate the possibility of providing HGS estimation by using foot-motion signals measured by an in-shoe motion sensor (IMS) embedded in an insole to achieve high precision HGS assessment in daily living. The foot-motion signals were collected from 62 elder participants (27 men and 35 women). Their HGSs were assessed by a hand dynamometer. Gait parameters, individual properties, and predictors derived from foot-motion signal features in one gait cycle were selected as candidates. Statistical parametric mapping analyses were used to generate predictors from the foot-motion signals. Prior to estimation construction, least absolute shrinkage and selection operator was applied to reduce redundant predictors from candidates. Linear regression models for HGS estimation of men and women were constructed. As the results, we discovered new effective predictors for HGS estimation from foot motions and successfully constructed HGS estimation models that achieved "excellent" agreement with the reference according to intra-class coefficients, and mean absolute errors of 2.96 and 2.57 kg for men and women in leave-one-subject-out cross-validation, respectively. These results suggest that HGS can be estimated with high precision by IMS-measured foot motion and more effective frailty identification in daily living is possible through wearing an IMS.

Method for Estimating Temporal Gait Parameters Concerning Bilateral Lower Limbs of Healthy Subjects Using a Single In-Shoe Motion Sensor through a Gait Event Detection Approach.

To expand the potential use of in-shoe motion sensors (IMSs) in daily healthcare or activity monitoring applications for healthy subjects, we propose a real-time temporal estimation method for gait parameters concerning bilateral lower limbs (GPBLLs) that uses a single IMS and is based on a gait event detection approach. To validate the established methods, data from 26 participants recorded by an IMS and a reference 3D motion analysis system were compared. The agreement between the proposed method and the reference system was evaluated by the intraclass correlation coefficient (ICC). The results showed that, by averaging over five continuous effective strides, all time parameters achieved precisions of no more than 30 ms and agreement at the "excellent" level, and the symmetry indexes of the stride time and stance phase time achieved precisions of 1.0% and 3.0%, respectively, and agreement at the "good" level. These results suggest our method is effective and shows promise for wide use in many daily healthcare or activity monitoring applications for healthy subjects.

An algorithm for real time minimum toe clearance estimation from signal of in-shoe motion sensor.

An algorithm has been constructed for estimating minimum toe clearance (MTC), an important gait parameter previously proven to be a critical indicator of tripping risk. It uses data from a previously reported in-shoe motion sensor (IMS) for detecting gait events. First, candidate feature points in the IMS signal for use in detecting MTC events were identified. Then, the temporal agreement between each feature point and target MTC event was evaluated. Next, the accuracy and precision of the MTC estimated using each feature point was evaluated using a reference value obtained using a 3-D optical motion-capture system. The MTC was estimated using a geometric model and the IMS signal corresponding to the predicted MTC event. Once the best candidate feature point was identified, a real-time MTC estimation algorithm for use with an IMS was constructed. The mean values and standard deviations of measured foot motions obtained in a previous study were used for evaluating accuracy and precision. The results suggest that MTC events can be estimated by detecting the crossing point between the acceleration waveforms in the anterior-posterior and superior-inferior directions in an accuracy of 2.0% gait cycle. Using this feature point enables the MTC to be estimated in real time with an accuracy of 8.6 mm, which will enable monitoring of MTC in daily living.

Ankle-foot orthosis with dorsiflexion resistance using spring-cam mechanism increases knee flexion in the swing phase during walking in stroke patients with hemiplegia.

BACKGROUND: Ankle-foot orthoses with plantarflexion resistance (AFO-Ps) improve knee flexion in the stance phase on the paretic side in patients with hemiparesis. However, AFO-Ps decrease ankle power generation in the late stance phase and do not improve the knee flexion in the swing phase based on insufficient push-off at the late stance, resulting in lower toe clearance. RESEARCH QUESTION: This study sought to investigate the effect of an AFO with dorsiflexion resistance, which was implemented by our developed device with spring-cam mechanism attached to the AFO-P (Gait Solution; Pacific Supply Co., Ltd., Japan), on kinetics and kinematics in the lower limb during gait in patients with hemiparesis. METHODS: Eleven patients with hemiparesis due to stroke walked on a 7-m walkway at a self-selected comfortable pace in the following conditions: (a) walking using the AFO-P with the proposed device with a spring-cam mechanism (AFO-PCAM), (b) walking using the AFO-P without our device (AFO-P), and (c) walking using no device (barefoot condition). Gait kinematics and kinetics were collected using a three-dimensional motion analysis system and four ground-reaction force plates. Changes in all parameters from the barefoot to AFO-PCAM and AFO-P conditions were compared using the Wilcoxon signed-rank test. RESULTS: In the AFO-PCAM condition, decrease in the maximum ankle power generation in the late-stance phase was significantly smaller than that in the AFO-P condition (p = 0.041). We noted a significant higher change in knee flexion in the paretic swing phase in the AFO-PCAM condition relative to that in the AFO-P condition (p = 0.016). The effect size for the comparisons of change was large (r ≧ 0.5). SIGNIFICANCE: Our device facilitated the realization of the ankle plantarflexion power in the late-stance phase because of dorsiflexion resistance, increasing the knee flexion angle during the swing phase.

Three-dimensional GRF and CoP Estimation during Stair and Slope Ascent/Descent with Wearable IMUs and Foot Pressure Sensors.

Wearable systems for gait analysis in daily living have been recently developed. Previous studies have demonstrated the significant potential of these systems; however, most of them focused on the level-walking condition, which is a limited portion of daily activities. To provide a new contribution to the gait analysis field, we have developed the first models for estimating three-dimensional (3D) ground reaction force (GRF) and center of pressure (CoP) during stair and slope ascent/descent with wearable sensors. Our system comprises light weight inertial measurement units (IMUs) and foot pressure sensors. We modeled the correlation between the measurements obtained with the wearable sensors and the ground truth of GRF/CoP from force plates, on the basis of linear regression models. Twenty healthy subjects completed a collection of ascent/descent tasks on stairs or slopes. We tested our models using cross-validation to evaluate the estimation accuracy in terms of the root mean square error (RMSE), the normalized RMSE (NRMSE), and the Pearson's correlation coefficient between the estimated GRF/CoP and those obtained from force plates. The experimental results showed practical estimation accuracy was obtained for GRF (RMSE ≤ 44.94 N) and CoP (RMSE ≤19.43 mm). Our system promises to contribute to clinical and sports medicine research by serving as a novel tool for assessing stair and slope ascent/descent outside laboratory environments.