Huiying Yu

Application of wearable sensors for human gait analysis using fuzzy computational algorithm

Abstract The authors have developed and tested a wearable inertial sensor system for the acquisition of gait features. The sensors were placed on anatomical segments of the lower limb: foot, shank, thigh, and hip, and the motion data were then captured in conjunction with 3D ground reaction forces (GRFs). The method of relational matrix was applied to develop a rule-based system, an intelligent fuzzy computational algorithm. The rule-based system provides a feature matrix model representing the strength of association or interaction amongst the elements of the gait functions (limb-segments accelerations and GRFs) throughout the gait cycle. A comparison between the reference rule-based data and an input test data was evaluated using a fuzzy similarity algorithm. This system was tested and evaluated using two subject groups: 10 healthy subjects were recruited to establish the reference fuzzy rule-base, and 4 relapsing remitting multiple sclerosis subjects were used as an input test data; and the grade of similarity between them was evaluated. This similarity provides a quantitative assessment of mobility state of the impaired subject. This algorithmic tool may be helpful to the clinician in the identification of pathological gait impairments, prescribe treatment, and assess the improvements in response to therapeutic intervention.

Automatic classification of pathological gait patterns using ground reaction forces and machine learning algorithms

An automated gait classification method is developed in this study, which can be applied to analysis and to classify pathological gait patterns using 3D ground reaction force (GRFs) data. The study involved the discrimination of gait patterns of healthy, cerebral palsy (CP) and multiple sclerosis subjects. The acquired 3D GRFs data were categorized into three groups. Two different algorithms were used to extract the gait features; the GRFs parameters and the discrete wavelet transform (DWT), respectively. Nearest neighbor classifier (NNC) and artificial neural networks (ANN) were also investigated for the classification of gait features in this study. Furthermore, different feature sets were formed using a combination of the 3D GRFs components (mediolateral, anterioposterior, and vertical) and their various impacts on the acquired results were evaluated. The best leave-one-out (LOO) classification accuracy 85% was achieved. The results showed some improvement through the application of a features selection algorithm based on M-shaped value of vertical force and the statistical test ANOVA of mediolateral and anterioposterior forces. The optimal feature set of six features enhanced the accuracy to 95%. This work can provide an automated gait classification tool that may be useful to the clinician in the diagnosis and identification of pathological gait impairments.

Gait Variability while walking with three different speeds

Gait Variability is defined as changes in gait parameters from one stride to the next. Gait variability increases in individuals affected by neurodegenerative conditions such as Parkinsons disease and Huntington disease, and also with falls in the elderly and incident mobility disability. In this work, we study speed-related and age-related gait variabilities in healthy adults. Ten participants, five females (three young and two middle-aged) and five males (three young and two middle-aged) were recruited to walk on an instrumented treadmill for three minutes in this study. The gait variables (stride length, stride width, stride time and stride velocity) were extracted and processed from camera motion system. Results: slow speed walking increased gait variability with all gait variables; only stride width variability was increased significantly in middle aged subjects compared with the young subjects (p≪0.05), there were no changes in other variables. Based on gait variability differences in age, height and body mass, we propose to design a knowledge-base membership function of gait variability for all the related gait variables with different heights and weights (BMI) as cofactors in each age group. An automatic diagnostic tool, Fuzzy Inferential Reasoning system, will help the clinician to identify pathological impairment from normal.

Identification of Human Gait using Fuzzy Inferential Reasoning

The restoration of healthy locomotion (gait) after stroke, traumatic brain injury, and spinal cord injury, is a major task in neurological rehabilitation. The rehabilitation process is labor intensive. Patient evaluation is often subjective, foiling determination of precise rehabilitation goals and assessment of treatment effects. To date it is the experienced clinician who continues to perform functional gait assessment and training in the absence of virtually any technological assistance. This paper introduces an algorithm capable of identifying human gait patterns. The fuzzy inferential reasoning uses typical joint angle trajectories to identify varying gait patterns. The algorithm will, thus, offer doctors, therapists, and patients a significant tool to assess the efficacy and outcomes of medical rehabilitation therapies and practices.

Application of fuzzy sets for assisting the physician's model of functional impairments in human locomotion

The authors have developed a system to assist clinicians reliably assess, at an early post-insult stage, the degree of disability the patient will ultimately experience. Physician decision processes offered to date, especially those relative to diagnosis and patient treatment, suffer from the inability to incorporate all useful data on the patient. We present a computational intelligence algorithm based on fuzzy clustering the theory of fuzzy sets and systems techniques to aid the physician to evaluate the complete representation of information emanating from the measured kinetic, kinematics and electromyographic data from the patient. The fuzzy clustering technique helps develop membership functions as an optimization task. The calculated membership grades are organized in the form of optimized partition matrix. As the optimization method operates on available data, it attempts to reflect their characteristics in the resulting constructs, e.g. a distribution of the prototypical values of the clusters.

Recognition and decision-making algorithm in human locomotion based on the principles of fuzzy reasoning

The authors introduce a fuzzy rule-based algorithm to evaluate the activation patterns of muscles of the lower limb with respect to the gait phases during normal human walking. A relational matrix was established as a Cartesian product between the activation behaviors of muscles of the lower limb within the seven gait phases during normal walking. This relational matrix is an expression of the strength of association between the muscles and the gait phases. The resulting knowledge-base, therefore, depicts the relationship between the muscles in the respective gait phases during normal walking tasks. The cross-correlation between an input relational matrix and the knowledge base will provide a diagnostic assessment of the neurological state of the subject.

Application of wearable miniature non-invasive sensory system in human locomotion using soft computing algorithm

The authors have designed and tested a wearable miniature noninvasive sensory system for the acquisition of gait features. The sensors are placed on anatomical segments of the lower limb, and motion data was then acquired in conjunction with electromyography (EMG) for muscle activities, and instrumented treadmill for ground reaction forces (GRF). A relational matrix was established between the limb-segment accelerations and the gait phases. A further relational matrix was established between the EMG data and the gait phases. With these pieces of information, a fuzzy rule-based system was established. This rule-based system depicts the strength of association or interaction between limb-segments accelerations, EMG, and gait phases. The outcome of measurements between the rule-based data and the randomized input data were evaluated using a fuzzy similarity algorithm. This algorithm offers the possibility to perform functional comparisons using different sources of information. It can provide a quantitative assessment of gait function.