Gait Disorder Detection and Classification Method Using Inertia Measurement Unit for Augmented Feedback Training in Wearable Devices

Parkinson’s disease (PD) is a common neurodegenerative disease, one of the symptoms of which is a gait disorder, which decreases gait speed and cadence. Recently, augmented feedback training has been considered to achieve effective physical rehabilitation. Therefore, we have devised a numerical modeling process and algorithm for gait detection and classification (GDC) that actively utilizes augmented feedback training. The numerical model converted each joint angle into a magnitude of acceleration (MoA) and a Z-axis angular velocity (ZAV) parameter. Subsequently, we confirmed the validity of both the GDC numerical modeling and algorithm. As a result, a higher gait detection and classification rate (GDCR) could be observed at a higher gait speed and lower acceleration threshold (AT) and gyroscopic threshold (GT). However, the pattern of the GDCR was ambiguous if the patient was affected by a gait disorder compared to a normal user. To utilize the relationships between the GDCR, AT, GT, and gait speed, we controlled the GDCR by using AT and GT as inputs, which we found to be a reasonable methodology. Moreover, the GDC algorithm could distinguish between normal people and people who suffered from gait disorders. Consequently, the GDC method could be used for rehabilitation and gait evaluation.

Using Remote Monitoring and Machine Learning to Classify Slam Events of Wave Piercing Catamarans

An onboard monitoring system can measure features such as stress cycles counts and provide warnings due to slamming. Considering current technology trends there is the opportunity of incorporating machine learning methods into monitoring systems. A hull monitoring system has been developed and installed on a 111 m wave piercing catamaran (Hull 091) to remotely monitor the ship kinematics and hull structural responses. Parallel to that, an existing dataset of a similar vessel (Hull 061) was analysed using unsupervised and supervised learning models; these were found to be beneficial for the classification of bow entry events according to key kinematic parameters. A comparison of different algorithms including linear support vector machines, naïve Bayes and decision tree for the bow entry classification were conducted. In addition, using empirical probability distributions, the likelihood of wet-deck slamming was estimated given a vertical bow acceleration threshold of 1 in head seas, clustering the feature space with the approximate probabilities of 0.001, 0.030 and 0.25.

Measuring threshold and latency of motion perception on a swinging bed

Introduction Our objective was to develop and to evaluate a system to measure latency and threshold of pendular motion perception based on a swinging bed. Materials and methods This prospective study included 30 healthy adults (age: 32 ± 12 years). All subjects were tested twice with a 10 min. interval. A second trial was conducted 2 to 15 days after. A rehabilitation swinging bed was connected to an electronic device emitting a beep at the beginning of each oscillation phase with an adjustable time lag. Subjects were blindfolded and auditory cues other than the beep were minimized. The acceleration threshold was measured by letting the bed oscillate freely until a natural break and asking the patient when he did not perceive any motion. The perception latency was determined by asking the patient to indicate whether the beep and the peak of each oscillation were synchronous. The time lag between sound and peak of the head position was swept from -750 to +750 ms by 50 ms increments. Results The mean acceleration threshold was 9.2±4.60 cm/s2. The range width of the synchronous perception interval was estimated as 535±190 ms. The point of subjective synchronicity defined as the center of this interval was -195±106 ms (n = 30). The test-retest evaluation in the same trial showed an acceptable reproducibility for the acceleration threshold and good to excellent for all parameters related to sound-movement latency. Conclusion Swinging bed combined to sound stimulation can provide reproducible information on movement perception in a simple and non-invasive manner with highly reproducible results.

High-Acceleration Movement, Muscle Damage, and Perceived Exertion in Basketball Games

Purpose: To evaluate the effect of the number of high-acceleration movements on muscle damage and the rating of perceived exertion (RPE) in basketball games. Methods: Twenty-one male collegiate basketball players (mean age, 20.0 [1.0] y) were included. A triaxial accelerometer was used to measure acceleration in basketball-simulated scrimmages. To detect higher physical load during the actual game, the resultant acceleration was calculated, and 3 thresholds were set: >4G, >6G, and >8G resultant accelerations. The number of the extracted movements was calculated at each acceleration threshold. Plasma creatine kinase (CK) levels (marker of muscle damage) were estimated before and 24 hours after the match, and the session-RPE load was calculated within 30 minutes after the match. Pearson product-moment correlations with 95% confidence intervals were used to determine the relationships between the number of high-acceleration movements and plasma CK and session-RPE load. Results: Significant correlations were observed between the number of high-acceleration movements >8G and CK level (r = .74; 95% confidence interval, 0.44–0.89; P < .0001). Furthermore, the correlation coefficient between acceleration and CK increased with increased acceleration threshold (>4G: r = .65; >6G: r = .69). Contrastingly, the correlation coefficient between acceleration and the session-RPE load decreased with increased acceleration threshold (>4G: r = .72; >6G: r = .52; >8G: r = .43). Conclusions: The session-RPE reflects the total amount of movement, while the high-acceleration movement reflects the momentary large impact load or intensity, and they evaluated different factors. Basketball coaching and conditioning professionals recommended combining acceleration and session-RPE when monitoring the load of athletes.

Water drops bouncing off vertically vibrating textured surfaces

We investigate the conditions that determine the detachment of a water drop from different vibrating textured plates by using vertical vibrations. The plate surfaces were patterned by a lattice of pillars of different shapes with different geometrical arrangements. The acceleration threshold for the water droplet to bounce off the surfaces was measured as a function of the excitation frequency. In each case, the acceleration threshold presents a minimum at the natural frequency of the droplet. The minimum acceleration required for the take-off is larger for small droplets than for large droplets. Namely, one finds that the value of the threshold depends on the size of the droplet and on the maximum apparent contact area between the droplet and the substrate. The theoretical model takes into account the energy necessary to break the capillary bridges between the droplet and the pillars of the surface. This model captures the main ingredients explaining the drop size dependence of the acceleration threshold for the take-off.

The Dimensionless Squared Jerk: An Objective Parameter That Improves Assessment of Hand Motion Analysis during Simulated Shoulder Arthroscopy

Purpose. Attempts to quantify hand movements of surgeons during arthroscopic surgery faced limited progress beyond motion analysis of hands and/or instruments. Surrogate markers such as procedure time have been used. The dimensionless squared jerk (DSJ) is a measure of deliberate hand movements. This study tests the ability of DSJ to differentiate novice and expert surgeons (construct validity) whilst performing simulated arthroscopic shoulder surgical tasks. Methods. Six residents (novice group) and six consultants (expert group) participated in this study. Participants performed three validated tasks sequentially under the same experimental setup (one performance). Each participant had ten performances assessed. Hand movements were recorded with optical tracking system. The DSJ, time taken, total path length, multiple measures of acceleration, and number of movements were recorded. Results. There were significant differences between novices and experts when assessed using time, number of movements with average and minimal acceleration threshold, and DSJ. No significant differences were observed in maximum acceleration, total path length, and number of movements with 10m/s2 acceleration threshold. Conclusion. DSJ is an objective parameter that can differentiate novice and expert surgeons’ simulated arthroscopic performances. We propose DSJ as an adjunct to more conventional parameters for arthroscopic surgery skills assessment.