Laser-light cueing shoes with integrated foot pressure and inertial sensing for investigating the impact of visual cueing on gait characteristics in Parkinson's disease individuals

Gait disorders are a fundamental challenge in Parkinson's disease (PD). The use of laser-light visual cues emitted from shoes has demonstrated effective in improving freezing of gait within less restrictive environments. However, the effectiveness of shoes-based laser-light cueing may vary among individuals with PD who have different types of impairments. We introduced an innovative laser-light visual shoes system capable of producing alternating visual cues for the left and right feet through one-side cueing at a time, while simultaneously recording foot inertial data and foot pressures. The effects of this visual cueing system on gait patterns were assessed in individuals with PD, both those with well-gait and those with worse-gait. Our device successfully quantified gait characteristics, including the asymmetry in the center of pressure trajectory, in individuals with PD. Furthermore, visual cueing prolonged stride times and increased the percentage of stance phase, while concurrently reducing stride length in PD individuals with well-gait. Conversely, in PD individuals with worse-gait, visual cueing resulted in a decreased freeze index and a reduction in the proportion of intervals prone to freezing episodes. The effects of visual cueing varied between PD individuals with well-gait and those with worse-gait. Visual cueing slowed down gait in the well-gait group while it appeared to mitigate freezing episodes in worse-gait group. Future researches, including enhancements to extend the projection distance of visual cues and clinical assessments conducted in real-world settings, will help establish the clinical utility of our proposed visual cueing system.

Single-Element Dual-Interferometer for Precision Inertial Sensing: Sub-Picometer Structural Stability and Performance as a Reference for Laser Frequency Stabilization

Future GRACE-like geodesy missions could benefit from adopting accelerometer technology akin to that of the LISA Pathfinder, which employed laser interferometric readout at the sub-picometer level in addition to the conventional capacitive sensing, which is at best at the level of 100 pm. Improving accelerometer performance holds great potential to enhance the scientific output of forthcoming missions, carrying invaluable implications for research in climate, water resource management, and disaster risk reduction. To reach sub-picometer displacement sensing precision in the millihertz range, laser interferometers rely on suppression of laser-frequency noise by several orders of magnitude. Many optical frequency stabilization methods are available with varying levels of complexity, size, and performance. In this paper, we describe the performance of a Mach-Zehnder interferometer based on a compact monolithic optic. The setup consists of a commercial fiber injector, a custom-designed pentaprism used to split and recombine the laser beam, and two photoreceivers placed at the complementary output ports of the interferometer. The structural stability of the prism is transferred to the laser frequency via amplification, integration, and feedback of the balanced-detection signal, achieving a fractional frequency instability better than 6 parts in 1013, corresponding to an interferometer pathlength stability better than 1pm/Hz. The prism was designed to host a second interferometer to interrogate the position of a test mass. This optical scheme has been dubbed "single-element dual-interferometer" or SEDI.

Deep Neural Network for the Detections of Fall and Physical Activities Using Foot Pressures and Inertial Sensing

Fall detection and physical activity (PA) classification are important health maintenance issues for the elderly and people with mobility dysfunctions. The literature review showed that most studies concerning fall detection and PA classification addressed these issues individually, and many were based on inertial sensing from the trunk and upper extremities. While shoes are common footwear in daily off-bed activities, most of the aforementioned studies did not focus much on shoe-based measurements. In this paper, we propose a novel footwear approach to detect falls and classify various types of PAs based on a convolutional neural network and recurrent neural network hybrid. The footwear-based detections using deep-learning technology were demonstrated to be efficient based on the data collected from 32 participants, each performing simulated falls and various types of PAs: fall detection with inertial measures had a higher F1-score than detection using foot pressures; the detections of dynamic PAs (jump, jog, walks) had higher F1-scores while using inertial measures, whereas the detections of static PAs (sit, stand) had higher F1-scores while using foot pressures; the combination of foot pressures and inertial measures was most efficient in detecting fall, static, and dynamic PAs.

Adaptive Inertial Sensor-Based Step Length Estimation Model

Pedestrian dead reckoning (PDR) using inertial sensors has paved the way for developing several approaches to step length estimation. In particular, emerging step length estimation models are readily available to be utilized on smartphones, yet they are seldom formulated considering the kinematics of the human body during walking in combination with measured step lengths. We present a new step length estimation model based on the acceleration magnitude and step frequency inputs herein. Spatial positions of anatomical landmarks on the human body during walking, tracked by an optical measurement system, were utilized in the derivation process. We evaluated the performance of the proposed model using our publicly available dataset that includes measurements collected for two types of walking modes, i.e., walking on a treadmill and rectangular-shaped test polygon. The proposed model achieved an overall mean absolute error (MAE) of 5.64 cm on the treadmill and an overall mean walked distance error of 4.55% on the test polygon, outperforming all the models selected for the comparison. The proposed model was also least affected by walking speed and is unaffected by smartphone orientation. Due to its promising results and favorable characteristics, it could present an appealing alternative for step length estimation in PDR-based approaches.

Identification of Movements and Postures Using Wearable Sensors for Implementation in a Bi-Hormonal Artificial Pancreas System

BACKGROUND

Closed loop bi-hormonal artificial pancreas systems, such as the artificial pancreas (AP™) developed by Inreda Diabetic B.V., control blood glucose levels of type 1 diabetes mellitus patients via closed loop regulation. As the AP™ currently does not classify postures and movements to estimate metabolic energy consumption to correct hormone administration levels, considerable improvements to the system can be made. Therefore, this research aimed to investigate the possibility to use the current system to identify several postures and movements.

METHODS

seven healthy participants took part in an experiment where sequences of postures and movements were performed to train and assess a computationally sparing algorithm.

RESULTS

Using accelerometers, one on the hip and two on the abdomen, user-specific models achieved classification accuracies of 86.5% using only the hip sensor and 87.3% when including the abdomen sensors. With additional accelerometers on the sternum and upper leg for identification, 90.0% of the classified postures and movements were correct.

CONCLUSIONS

The current hardware configuration of the AP™ poses no limitation to the identification of postures and movements. If future research shows that identification can still be done accurately in a daily life setting, this algorithm may be an improvement for the AP™ to sense physical activity.

Investigation and Mitigation of Noise Contributions in a Compact Heterodyne Interferometer

We present a noise estimation and subtraction algorithm capable of increasing the sensitivity of heterodyne laser interferometers by one order of magnitude. The heterodyne interferometer is specially designed for dynamic measurements of a test mass in the application of sub-Hz inertial sensing. A noise floor of 3.31×10−11m/Hz at 100 mHz is achieved after applying our noise subtraction algorithm to a benchtop prototype interferometer that showed a noise level of 2.76×10−10m/Hz at 100 mHz when tested in vacuum at levels of 3×10−5 Torr. Based on the previous results, we investigated noise estimation and subtraction techniques of non-linear optical pathlength noise, laser frequency noise, and temperature fluctuations in heterodyne laser interferometers. For each noise source, we identified its contribution and removed it from the measurement by linear fitting or a spectral analysis algorithm. The noise correction algorithm we present in this article can be generally applied to heterodyne laser interferometers.

Single-Element Dual-Interferometer for Precision Inertial Sensing

Tracking moving masses in several degrees of freedom with high precision and large dynamic range is a central aspect in many current and future gravitational physics experiments. Laser interferometers have been established as one of the tools of choice for such measurement schemes. Using sinusoidal phase modulation homodyne interferometry allows a drastic reduction of the complexity of the optical setup, a key limitation of multi-channel interferometry. By shifting the complexity of the setup to the signal processing stage, these methods enable devices with a size and weight not feasible using conventional techniques. In this paper we present the design of a novel sensor topology based on deep frequency modulation interferometry: the self-referenced single-element dual-interferometer (SEDI) inertial sensor, which takes simplification one step further by accommodating two interferometers in one optic. Using a combination of computer models and analytical methods we show that an inertial sensor with sub-picometer precision for frequencies above 10 mHz, in a package of a few cubic inches, seems feasible with our approach. Moreover we show that by combining two of these devices it is possible to reach sub-picometer precision down to 2 mHz. In combination with the given compactness, this makes the SEDI sensor a promising approach for applications in high precision inertial sensing for both next-generation space-based gravity missions employing drag-free control, and ground-based experiments employing inertial isolation systems with optical readout.

Detection of the intention to grasp during reach movements

INTRODUCTION

Soft-robotic gloves have been developed to enhance grip to support stroke patients during daily life tasks. Studies showed that users perform tasks faster without the glove as compared to with the glove. It was investigated whether it is possible to detect grasp intention earlier than using force sensors to enhance the performance of the glove.

METHODS

This was studied by distinguishing reach-to-grasp movements from reach movements without the intention to grasp, using minimal inertial sensing and machine learning. Both single-user and multi-user support vector machine classifiers were investigated. Data were gathered during an experiment with healthy subjects, in which they were asked to perform grasp and reach movements.

RESULTS

Experimental results show a mean accuracy of 98.2% for single-user and of 91.4% for multi-user classification, both using only two sensors: one on the hand and one on the middle finger. Furthermore, it was found that using only 40% of the trial length, an accuracy of 85.3% was achieved, which would allow for an earlier prediction of grasp during the reach movement by 1200 ms.

CONCLUSIONS

Based on these promising results, further research will be done to investigate the possibility to use classification of the movements in stroke patients.

Usability Evaluations of a Wearable Inertial Sensing System and Quality of Movement Metrics for Stroke Survivors by Care Professionals

Background

Inertial motion capture systems are used in many applications such as measuring the movement quality in stroke survivors. The absence of clinical effectiveness and usability evidence in these assistive technologies into rehabilitation has delayed the transition of research into clinical practice. Recently, a new inertial motion capture system was developed in a project, called INTERACTION, to objectively measure the quality of movement (QoM) in stroke survivors during daily-life activity. With INTERACTION, we are to be able to investigate into what happens with patients after discharge from the hospital. Resulting QoM metrics, where a metric is defined as a measure of some property, are subsequently presented to care professionals. Metrics include for example: reaching distance, walking speed, and hand distribution plots. The latter shows a density plot of the hand position in the transversal plane. The objective of this study is to investigate the opinions of care professionals in using these metrics obtained from INTERACTION and its usability.

Methods

By means of a semi-structured interview, guided by a presentation, presenting two patient reports. Each report includes several QoM metric (like reaching distance, hand position density plots, shoulder abduction) results obtained during daily-life measurements and in clinic and were evaluated by care professionals not related to the project. The results were compared with care professionals involved within the INTERACTION project. Furthermore, two questionnaires (5-point Likert and open questionnaire) were handed over to rate the usability of the metrics and to investigate if they would like such a system in their clinic.

Results

Eleven interviews were conducted, where each interview included either two or three care professionals as a group, in Switzerland and The Netherlands. Evaluation of the case reports (CRs) by participants and INTERACTION members showed a high correlation for both lower and upper extremity metrics. Participants were most in favor of hand distribution plots during daily-life activities. All participants mentioned that visualizing QoM of stroke survivors over time during daily-life activities has more possibilities compared to current clinical assessments. They also mentioned that these metrics could be important for self-evaluation of stroke survivors.

Discussion

The results showed that most participants were able to understand the metrics presented in the CRs. For a few metrics, it remained difficult to assess the underlying cause of the QoM. Hence, a combination of metrics is needed to get a better insight of the patient. Furthermore, it remains important to report the state (e.g., how the patient feels), its surroundings (outside, inside the house, on a slippery surface), and detail of specific activities (does the patient grasps a piece of paper or a heavy cooking pan but also dual tasks). Altogether, it remains a questions how to determine what the patient is doing and where the patient is doing his or her activities.

Usability Evaluation of a VibroTactile Feedback System in Stroke Subjects

Background

To increase the functional capabilities of stroke subjects during activities of daily living, patients receive rehabilitative training to recover adequate motor control. With the goal to motivate self-training by use of the arm in daily life tasks, a sensor system (Arm Usage Coach, AUC) was developed that provides VibroTactile (VT) feedback if the patient does not move the affected arm above a certain threshold level. The objective of this study is to investigate the usability of this system in stroke subjects.

Method

The study was designed as a usability and user acceptance study of feedback modalities. Stroke subjects with mild to moderate arm impairments were enrolled. The subjects wore two AUC devices one on each wrist. VT feedback was given by the device on the affected arm. A semi-structured interview was performed before and after a measurement session with the AUC. In addition, the System Usability Scale (SUS) questionnaire was given.

Results

Ten ischemic chronic stroke patients (39 ± 38 months after stroke) were recruited. Four out of 10 subjects have worn the VT feedback on their dominant, affected arm. In the pre-measurement interview, eight participants indicated a preference for acoustic or visual over VT feedback. In the post evaluation interview, nine of 10 participants preferred VT over visual and acoustic feedback. On average, the AUC gave VT feedback six times during the measurement session. All participants, with the exception of one, used their dominant arm more then the non-dominant. For the SUS, eight participants responded above 80%, one between 70 and 80%, and one participant responded below 50%.

Discussion

More patients accepted and valued VT feedback after the test period, hence VT is a feasible feedback modality. The AUC can be used as a telerehabilitation device to train and maintain upper extremity use in daily life tasks.

Objective Evaluation of the Quality of Movement in Daily Life after Stroke

Background

Stroke survivors are commonly left with disabilities that impair activities of daily living. The main objective of their rehabilitation program is to maximize the functional performance at home. However, the actual performance of patients in their home environment is unknown. Therefore, objective evaluation of daily life activities of stroke survivors in their physical interaction with the environment is essential for optimal guidance of rehabilitation therapy. Monitoring daily life movements could be very challenging, as it may result in large amounts of data, without any context. Therefore, suitable metrics are necessary to quantify relevant aspects of movement performance during daily life. The objective of this study is to develop data processing methods, which can be used to process movement data into relevant metrics for the evaluation of intra-patient differences in quality of movements in a daily life setting.

Methods

Based on an iterative requirement process, functional and technical requirements were formulated. These were prioritized resulting in a coherent set of metrics. An activity monitor was developed to give context to captured movement data at home. Finally, the metrics will be demonstrated in two stroke participants during and after their rehabilitation phases.

Results

By using the final set of metrics, quality of movement can be evaluated in a daily life setting. As example to demonstrate potential of presented methods, data of two stroke patients were successfully analyzed. Differences between in-clinic measurements and measurements during daily life are observed by applying the presented metrics and visualization methods. Heel height profiles show intra-patient differences in height, distance, stride profile, and variability between strides during a 10-m walk test in the clinic and walking at home. Differences in distance and stride profile between both feet were larger at home, than in clinic. For the upper extremities, the participant was able to reach further away from the pelvis and cover a larger area.

Discussion

Presented methods can be used for the objective evaluation of intra-patient differences in movement quality between in-clinic and daily life measurements. Any observed progression or deterioration of movement quality could be used to decide on continuing, stopping, or adjusting rehabilitation programs.

Activity recognition using a single accelerometer placed at the wrist or ankle

PURPOSE

Large physical activity surveillance projects such as the UK Biobank and NHANES are using wrist-worn accelerometer-based activity monitors that collect raw data. The goal is to increase wear time by asking subjects to wear the monitors on the wrist instead of the hip, and then to use information in the raw signal to improve activity type and intensity estimation. The purposes of this work was to obtain an algorithm to process wrist and ankle raw data and to classify behavior into four broad activity classes: ambulation, cycling, sedentary, and other activities.

METHODS

Participants (N = 33) wearing accelerometers on the wrist and ankle performed 26 daily activities. The accelerometer data were collected, cleaned, and preprocessed to extract features that characterize 2-, 4-, and 12.8-s data windows. Feature vectors encoding information about frequency and intensity of motion extracted from analysis of the raw signal were used with a support vector machine classifier to identify a subject's activity. Results were compared with categories classified by a human observer. Algorithms were validated using a leave-one-subject-out strategy. The computational complexity of each processing step was also evaluated.

RESULTS

With 12.8-s windows, the proposed strategy showed high classification accuracies for ankle data (95.0%) that decreased to 84.7% for wrist data. Shorter (4 s) windows only minimally decreased performances of the algorithm on the wrist to 84.2%.

CONCLUSIONS

A classification algorithm using 13 features shows good classification into the four classes given the complexity of the activities in the original data set. The algorithm is computationally efficient and could be implemented in real time on mobile devices with only 4-s latency.