Predicting lower limb joint kinematics using wearable motion sensors.

The aim of this study was to estimate sagittal plane ankle, knee and hip gait kinematics using 3D angular velocity and linear acceleration data from motion sensors on the foot and shank. We explored the accuracy of intra-subject predictions (i.e., where training and testing uses trials from the same subject) and inter-subject (where testing uses subjects different from the ones used for training) predictions, and the effect of loss of sensor data on prediction accuracy. Hip, knee and ankle kinematic data were collected using reflective markers. Simultaneously, foot and shank angular velocity and linear acceleration data were collected using small integrated accelerometers/gyroscope units. A generalised regression networks algorithm was used to predict the former from the latter. The best results were from intra-subject predictions, with very high correlations (0.93-0.99) and low mean absolute deviation (< or =2.3 degrees ) between measured kinematic joint angles and predicted angles. The inter-subject case produced poorer correlations (0.70-0.89) and larger absolute differences between measured and predicted angles, ranging from 4.91 degrees (left ankle) to 9.06 degrees (right hip). The angular velocity data added little to the accuracy of predictions and there was also minimal benefit to using sensor data from the shank. Thus, a wearable system based only on footwear mounted sensors and a simpler sensor set providing only acceleration data shows potential. Whilst predictions were generally stable when sensor data was lost, it remains to be seen whether the generalised regression networks algorithm is robust for other activities such as stair climbing.

The effect of the impedance of a thin hydrogel electrode on sensation during functional electrical stimulation.

Surface functional electrical stimulation results in stimulation of cutaneous receptors and discomfort. The degree of non-uniformity of current distribution in the cutaneous layers in the vicinity of the electrode may influence the sensation experienced. This paper describes the effects on sensation of a thin, high impedance electrode designed to reduce the non-uniformity of current distribution. Sensation associated with stimulation via a self-adhesive electrode with much higher impedance than conventional electrodes was compared with a low impedance electrode in a single-blinded, crossover study. The high impedance electrode does not alter either the current at which sensation is first registered. However, at higher currents, the high impedance electrode allows 9% more current to be passed for an equivalent sensation to that experienced with the conventional electrode. A 28% decrease in discomfort with the use of the high impedance electrode was also reported.

A generalisable methodology for stability assessment of walking aid users.

To assist balance and mobility, older adults are often prescribed walking aids. Nevertheless, surprisingly their use has been associated with increased falls-risk. To address this finding we first need to characterise a person's stability while using a walking aid. Therefore, we present a generalisable method for the assessment of stability of walking frame (WF) users. Our method, for the first time, considers user and device as a combined system. We define the combined centre of pressure (CoPsystem) of user and WF to be the point through which the resultant ground reaction force for all feet of both the WF and user acts if theresultant moment acts only around an axisperpendicular tothe ground plane. We also define the combined base of support (BoSsystem) to be the convex polygon formed by the boundaries of the anatomical and WF feet in contact with the ground and interconnecting lines between them. To measure these parameters we have developed an instrumented WF with a load cell in each foot which we use together with pressure-sensing insoles and a camera system, the latter providing the relative position of the WF and anatomical feet. Software uses the resulting data to calculate the stability margin of the combined system, defined as the distance between CoPsystem and the nearest edge of BoSsystem. Our software also calculates the weight supported through the frame and when each foot (of user and/or frame) is on the floor. Finally, we present experimental work demonstrating the value of our approach.

Effects of shoe sole geometry on toe clearance and walking stability in older adults.

Thirty-five percent of people above age 65 fall each year, and half of their falls are associated with tripping: tripping, an apparently 'mundane' everyday problem, therefore, significantly impacts on older people's health and associated medical costs. To avoid tripping and subsequent falling, sufficient toe clearance during the swing phase is crucial. We previously found that a rocker-shaped shoe sole enhances toe clearance in young adults, thereby decreasing their trip-risk. This study investigates whether such sole design also enhances older adults' toe clearance, without inadvertently affecting their walking stability. Toe clearance and its variability are reported together with measures of walking stability for twelve older adults, walking in shoes with rocker angles of 10°, 15°, and 20°. Surface inclinations (flat, incline, decline) were chosen to reflect a potential real-world environment. Toe clearance increased substantially from the 10° to the 15° rocker angle (p=0.003) without compromising measures of walking stability (p>0.05). A further increase in rocker angle to 20° resulted in less substantial enhancement of toe clearance and came at the cost of a decrease in gait speed on the decline. The novelty of this investigation lies in the exploration of the trade-off between reduction of trip-risk through footwear design and adverse effects on walking stability on real-life relevant surfaces. Our two studies suggest that the current focus on slip-resistance in footwear design may need to be generalised to include other factors that affect trip-risk.

Automatic detection of lift-off and touch-down of a pick-up walker using 3D kinematics.

Walking aids have been associated with falls and it is believed that incorrect use limits their usefulness. Measures are therefore needed that characterize their stable use and the classification of key events in walking aid movement is the first step in their development. This study presents an automated algorithm for detection of lift-off (LO) and touch-down (TD) events of a pick-up walker. For algorithm design and initial testing, a single user performed trials for which the four individual walker feet lifted off the ground and touched down again in various sequences, and for different amounts of frame loading (Dataset_1). For further validation, ten healthy young subjects walked with the pick-up walker on flat ground (Dataset_2a) and on a narrow beam (Dataset_2b), to challenge balance. One 88-year-old walking frame user was also assessed. Kinematic data were collected with a 3D optoelectronic camera system. The algorithm detected over 93% of events (Dataset_1), and 95% and 92% in Dataset_2a and b, respectively. Of the various LO/TD sequences, those associated with natural progression resulted in up to 100% correctly identified events. For the 88-year-old walking frame user, 96% of LO events and 93% of TD events were detected, demonstrating the potential of the approach.

Biomechanics for inclusive urban design: Effects of tactile paving on older adults' gait when crossing the street.

In light of our ageing population it is important that the urban environment is easily accessible and hence supports older adults' independence. Tactile 'blister' paving was originally designed to provide guidance for visually impaired people at pedestrian crossings. However, as research links irregular surfaces to falls in older adults, such paving may have an adverse effect on older people. We investigated the effects of tactile paving on older adults' gait in a scenario closely resembling crossing the street. Gait analysis of 32 healthy older adults showed that tactile, as compared to smooth, paving increases the variability in timing of foot placement by 20%, thereby indicating a disturbance of the rhythmic gait pattern. Moreover, toe clearance during the swing phase increased by 7% on tactile paving, and the ability to stop upon cue from the traffic light was compromised. These results need to be viewed under the consideration of limitations associated with laboratory studies and real world analysis is needed to fully understand their implications for urban design.

Effects of ramp negotiation, paving type and shoe sole geometry on toe clearance in young adults.

Trips are a major cause of falls and result from involuntary contact of the foot with the ground during the swing phase of gait. Adequate toe clearance during swing is therefore crucial for safe locomotion. To date, little is known about the effects of environmental factors and footwear on toe clearance. This study reports on modulation of toe clearance and toe clearance variability in response to changes in ground inclination, paving type, and shoe sole geometry. Toe clearance and toe clearance variability for ten healthy young adults were calculated two-fold: a) for the commonly-used position on the foremost part of the sole of the shoe and b) for the lowest of a total of 7 sole positions, located between the metatarsals and the toe tip across the entire width of the sole. Utilizing a full-factorial design we found that toe clearance was affected by ground inclination, paving type, and sole geometry regardless of the computational method used (with p-values<0.01) but the use of the foremost part of the sole for toe clearance calculation results is an overestimation of this value. Our findings highlight the importance of considering footwear and environmental factors when assessing the risk of tripping. Future work needs to investigate to which extent the same factors affect toe clearance in more vulnerable parts of the population.

A clinical set-up tool (CST) for rapid stimulator programming.

This paper presents an intuitive approach to the programming of Functional Electrical Stimulation (FES) hardware. A stimulation sequence is defined as a Finite State Machine (FSM) that is created and modified via an easy-to-use interface. Using a USB-equipped stimulator, data from an internal accelerometer are streamed to a PC whilst the subject undergoes the motion of interest. Simultaneously, the clinician labels the collected data via a number of key presses in order to train a sequence of simple classifiers. The 'optimal' (in a quantitative sense) parameters are then uploaded directly back to the stimulator for immediate stand-alone operation.

A dual-task approach to the evaluation of the myokinemetric signal as an alternative to EMG.

EMG is the signal widely used in neuromuscular control, biofeedback and measurement applications. Alternative physiological signals are available, but are used relatively infrequently. In the development of assistive devices, such as functional electrical stimulators, it is important to make the device as straightforward to use as possible. This is particularly relevant for patients with neurological and often associated cognitive impairments. Different physiological signals may require different degrees of attention to control, and advantage could be gained from selection of a signal that requires the least attention to control. However, relatively little work has been carried out on how to assess the demands of different physiological signals. This paper reports on the development of a novel experimental set up designed to address this problem and, in particular, to compare two different physiological signals, the EMG and the so-called MK signal. The paper presents the hardware design, including mechanical, electronic and software design, which involves data acquisition, parallel tasks and user-friendly interface. The system described could be adapted for evaluation of other physiological signals.