Freezing of gait (FoG) is a common gait deficit in advanced Parkinson's disease (PD). FoG events are associated with falls, interfere with daily life activities and impair quality of life. FoG is often resistant to pharmacologic treatment; therefore effective non-pharmacologic assistance is needed.We propose a wearable assistant, composed of a smartphone and wearable accelerometers, for online detection of FoG. The system is based on machine learning techniques for automatic detection of FoG episodes. When FoG is detected, the assistant provides rhythmic auditory cueing or vibrotactile feedback that stimulates the patient to resume walking.We tested our solution on more than 8h of recorded lab data from PD patients that experience FoG in daily life. We characterize the system performance on user-dependent and user-independent experiments, with respect to different machine learning algorithms, sensor placement and preprocessing window size. The final system was able to detect FoG events with an average sensitivity and specificity of more than 95%, and mean detection latency of 0.34s in user-dependent settings.

Gait impairments are among the most disabling symptoms in several musculoskeletal and neurological conditions, severely limiting personal autonomy. Wearable gait sensors have been attracting attention as diagnostic tool for gait and are emerging as promising tool for tutoring and guiding gait execution. If their popularity is continuously growing, still there is room for improvement, especially towards more accurate solutions for spatio-temporal gait parameters estimation. We present an implementation of a zero-velocity-update gait analysis system based on a Kalman filter and off-the-shelf shoe-worn inertial sensors. The algorithms for gait events and step length estimation were specifically designed to comply with pathological gait patterns. More so, an Android app was deployed to support fully wearable and stand-alone real-time gait analysis. Twelve healthy subjects were enrolled to preliminarily tune the algorithms; afterwards sixteen persons with Parkinson's disease were enrolled for a validation study. Over the 1314 strides collected on patients at three different speeds, the total root mean square difference on step length estimation between this system and a gold standard was 2.9%. This shows that the proposed method allows for an accurate gait analysis and paves the way to a new generation of mobile devices usable anywhere for monitoring and intervention.

During natural locomotion, the stiffness of the human knee is modulated continuously and subconsciously according to the demands of activity and terrain. Given modern actuator technology, powered transfemoral prostheses could theoretically provide a similar degree of sophistication and function. However, experimentally quantifying knee stiffness modulation during natural gait is challenging. Alternatively, joint stiffness could be estimated in a less disruptive manner using electromyography (EMG) combined with kinetic and kinematic measurements to estimate muscle force, together with models that relate muscle force to stiffness. Here we present the first step in that process, where we develop such an approach and evaluate it in isometric conditions, where experimental measurements are more feasible. Our EMG-guided modeling approach allows us to consider conditions with antagonistic muscle activation, a phenomenon commonly observed in physiological gait. Our validation shows that model-based estimates of knee joint stiffness coincide well with experimental data obtained using conventional perturbation techniques. We conclude that knee stiffness can be accurately estimated in isometric conditions without applying perturbations, which presents an important step towards our ultimate goal of quantifying knee stiffness during gait.

No abstract