Walking-speed estimation using a single inertial measurement unit for the older adults

A wearable sensor and machine learning estimate step length in older adults and patients with neurological disorders

Step length is an important diagnostic and prognostic measure of health and disease. Wearable devices can estimate step length continuously (e.g., in clinic or real-world settings), however, the accuracy of current estimation methods is not yet optimal. We developed machine-learning models to estimate step length based on data derived from a single lower-back inertial measurement unit worn by 472 young and older adults with different neurological conditions, including Parkinson's disease and healthy controls. Studying more than 80,000 steps, the best model showed high accuracy for a single step (root mean square error, RMSE = 6.08 cm, ICC(2,1) = 0.89) and higher accuracy when averaged over ten consecutive steps (RMSE = 4.79 cm, ICC(2,1) = 0.93), successfully reaching the predefined goal of an RMSE below 5 cm (often considered the minimal-clinically-important-difference). Combining machine-learning with a single, wearable sensor generates accurate step length measures, even in patients with neurologic disease. Additional research may be needed to further reduce the errors in certain conditions.

Beyond gait speed: exploring the added value of Inertial Measurement Unit-based measurements of gait in the estimation of the walking ability in daily life

BACKGROUND

Gait speed is often used to estimate the walking ability in daily life in people after stroke. While measuring gait with inertial measurement units (IMUs) during clinical assessment yields additional information, it remains unclear if this information can improve the estimation of the walking ability in daily life beyond gait speed.

OBJECTIVE

We evaluated the additive value of IMU-based gait features over a simple gait-speed measurement in the estimation of walking ability in people after stroke.

METHODS

Longitudinal data during clinical stroke rehabilitation were collected. The assessment consisted of two parts and was administered every three weeks. In the first part, participants walked for two minutes (2MWT) on a fourteen-meter path with three IMUs attached to low back and feet, from which multiple gait features, including gait speed, were calculated. The dimensionality of the corresponding gait features was reduced with a principal component analysis. In the second part, gait was measured for two consecutive days using one ankle-mounted IMU. Next, three measures of walking ability in daily life were calculated, including the number of steps per day, and the average and maximal gait speed. A gait-speed-only Linear Mixed Model was used to estimate the association between gait speed and each of the three measures of walking ability. Next, the principal components (PC), derived from the 2MWT, were added to the gait-speed-only model to evaluate if they were confounders or effect modifiers.

RESULTS

Eighty-one participants were measured during rehabilitation, resulting in 198 2MWTs and 135 corresponding walking-performance measurements. 106 Gait features were reduced to nine PCs with 85.1% explained variance. The linear mixed models demonstrated that gait speed was weakly associated with the average and maximum gait speed in daily life and moderately associated with the number of steps per day. The PCs did not considerably improve the outcomes in comparison to the gait speed only models.

CONCLUSIONS

Gait in people after stroke assessed in a clinical setting with IMUs differs from their walking ability in daily life. More research is needed to determine whether these discrepancies also occur in non-laboratory settings, and to identify additional non-gait factors that influence walking ability in daily life.

Mobilise-D insights to estimate real-world walking speed in multiple conditions with a wearable device

This study aimed to validate a wearable device's walking speed estimation pipeline, considering complexity, speed, and walking bout duration. The goal was to provide recommendations on the use of wearable devices for real-world mobility analysis. Participants with Parkinson's Disease, Multiple Sclerosis, Proximal Femoral Fracture, Chronic Obstructive Pulmonary Disease, Congestive Heart Failure, and healthy older adults (n = 97) were monitored in the laboratory and the real-world (2.5 h), using a lower back wearable device. Two walking speed estimation pipelines were validated across 4408/1298 (2.5 h/laboratory) detected walking bouts, compared to 4620/1365 bouts detected by a multi-sensor reference system. In the laboratory, the mean absolute error (MAE) and mean relative error (MRE) for walking speed estimation ranged from 0.06 to 0.12 m/s and - 2.1 to 14.4%, with ICCs (Intraclass correlation coefficients) between good (0.79) and excellent (0.91). Real-world MAE ranged from 0.09 to 0.13, MARE from 1.3 to 22.7%, with ICCs indicating moderate (0.57) to good (0.88) agreement. Lower errors were observed for cohorts without major gait impairments, less complex tasks, and longer walking bouts. The analytical pipelines demonstrated moderate to good accuracy in estimating walking speed. Accuracy depended on confounding factors, emphasizing the need for robust technical validation before clinical application.Trial registration: ISRCTN - 12246987.

Machine Learning Applications in Sarcopenia Detection and Management: A Comprehensive Survey

This extensive review examines sarcopenia, a condition characterized by a loss of muscle mass, stamina, and physical performance, with a particular emphasis on its detection and management using contemporary technologies. It highlights the lack of global agreement or standardization regarding the definition of sarcopenia and the various techniques used to measure muscle mass, stamina, and physical performance. The distinctive criteria employed by the European Working Group on Sarcopenia in Older People (EWGSOP) and the Asian Working Group for Sarcopenia (AWGSOP) for diagnosing sarcopenia are examined, emphasizing potential obstacles in comparing research results across studies. The paper delves into the use of machine learning techniques in sarcopenia detection and diagnosis, noting challenges such as data accessibility, data imbalance, and feature selection. It suggests that wearable devices, like activity trackers and smartwatches, could offer valuable insights into sarcopenia progression and aid individuals in monitoring and managing their condition. Additionally, the paper investigates the potential of blockchain technology and edge computing in healthcare data storage, discussing models and systems that leverage these technologies to secure patient data privacy and enhance personal health information management. However, it acknowledges the limitations of these models and systems, including inefficiencies in handling large volumes of medical data and the lack of dynamic selection capability. In conclusion, the paper provides a comprehensive summary of current sarcopenia research, emphasizing the potential of modern technologies in enhancing the detection and management of the condition while also highlighting the need for further research to address challenges in standardization, data management, and effective technology use.

A Comparison of Walking Behavior during the Instrumented TUG and Habitual Gait

The timed up and go test (TUG) is a common clinical functional balance test often used to complement findings on sensorimotor changes due to aging or sensory/motor dysfunction. The instrumented TUG can be used to obtain objective postural and gait measures that are more sensitive to mobility changes. We investigated whether gait and body coordination during TUG is representative of walking. We examined the walking phase of the TUG and compared gait metrics (stride duration and length, walking speed, and step frequency) and head/trunk accelerations to normal walking. The latter is a key aspect of postural control and can also reveal changes in sensory and motor function. Forty participants were recruited into three groups: young adults, older adults, and older adults with visual impairment. All performed the TUG and a short walking task wearing ultra-lightweight wireless IMUs on the head, chest, and right ankle. Gait and head/trunk acceleration metrics were comparable across tasks. Further, stride length and walking speed were correlated with the participants' age. Those with visual impairment walked significantly slower than sighted older adults. We suggest that the TUG can be a valuable tool for examining gait and stability during walking without the added time or space constraints.

Development of a Gait Feature-Based Model for Classifying Cognitive Disorders Using a Single Wearable Inertial Sensor

BACKGROUND AND OBJECTIVES

Gait changes are potential markers of cognitive disorders (CDs). We developed a model for classifying older adults with CD from those with normal cognition using gait speed and variability captured from a wearable inertial sensor and compared its diagnostic performance for CD with that of the model using the Mini-Mental State Examination (MMSE).

METHODS

We enrolled community-dwelling older adults with normal gait from the Korean Longitudinal Study on Cognitive Aging and Dementia and measured their gait features using a wearable inertial sensor placed at the center of body mass while they walked on a 14-m long walkway thrice at comfortable paces. We randomly split our entire dataset into the development (80%) and validation (20%) datasets. We developed a model for classifying CD using logistic regression analysis from the development dataset and validated it in the validation dataset. In both datasets, we compared the diagnostic performance of the model with that using the MMSE. We estimated optimal cutoff score of our model using receiver operator characteristic analysis.

RESULTS

In total, 595 participants were enrolled, of which 101 of them experienced CD. Our model included both gait speed and temporal gait variability and exhibited good diagnostic performance for classifying CD from normal cognition in both the development (area under the receiver operator characteristic curve [AUC] = 0.788, 95% CI 0.748-0.823, p < 0.001) and validation datasets (AUC = 0.811, 95% CI 0.729-0.877, p < 0.001). Our model showed comparable diagnostic performance for CD with that of the model using the MMSE in both the development (difference in AUC = 0.026, standard error [SE] = 0.043, z statistic = 0.610, p = 0.542) and validation datasets (difference in AUC = 0.070, SE = 0.073, z statistic = 0.956, p = 0.330). The optimal cutoff score of the gait-based model was >-1.56.

DISCUSSION

Our gait-based model using a wearable inertial sensor may be a promising diagnostic marker of CD in older adults.

CLASSIFICATION OF EVIDENCE

This study provides Class III evidence that gait analysis can accurately distinguish older adults with CDs from healthy controls.

Assessing real-world gait with digital technology? Validation, insights and recommendations from the Mobilise-D consortium

BACKGROUND

Although digital mobility outcomes (DMOs) can be readily calculated from real-world data collected with wearable devices and ad-hoc algorithms, technical validation is still required. The aim of this paper is to comparatively assess and validate DMOs estimated using real-world gait data from six different cohorts, focusing on gait sequence detection, foot initial contact detection (ICD), cadence (CAD) and stride length (SL) estimates.

METHODS

Twenty healthy older adults, 20 people with Parkinson's disease, 20 with multiple sclerosis, 19 with proximal femoral fracture, 17 with chronic obstructive pulmonary disease and 12 with congestive heart failure were monitored for 2.5 h in the real-world, using a single wearable device worn on the lower back. A reference system combining inertial modules with distance sensors and pressure insoles was used for comparison of DMOs from the single wearable device. We assessed and validated three algorithms for gait sequence detection, four for ICD, three for CAD and four for SL by concurrently comparing their performances (e.g., accuracy, specificity, sensitivity, absolute and relative errors). Additionally, the effects of walking bout (WB) speed and duration on algorithm performance were investigated.

RESULTS

We identified two cohort-specific top performing algorithms for gait sequence detection and CAD, and a single best for ICD and SL. Best gait sequence detection algorithms showed good performances (sensitivity > 0.73, positive predictive values > 0.75, specificity > 0.95, accuracy > 0.94). ICD and CAD algorithms presented excellent results, with sensitivity > 0.79, positive predictive values > 0.89 and relative errors < 11% for ICD and < 8.5% for CAD. The best identified SL algorithm showed lower performances than other DMOs (absolute error < 0.21 m). Lower performances across all DMOs were found for the cohort with most severe gait impairments (proximal femoral fracture). Algorithms' performances were lower for short walking bouts; slower gait speeds (< 0.5 m/s) resulted in reduced performance of the CAD and SL algorithms.

CONCLUSIONS

Overall, the identified algorithms enabled a robust estimation of key DMOs. Our findings showed that the choice of algorithm for estimation of gait sequence detection and CAD should be cohort-specific (e.g., slow walkers and with gait impairments). Short walking bout length and slow walking speed worsened algorithms' performances. Trial registration ISRCTN - 12246987.

A normative study of the gait features measured by a wearable inertia sensor in a healthy old population

BACKGROUND

Gait function impairments are associated with the risk of various medical conditions in older adults. As gait function declines with advancing age, normative data are required for proper interpretation of gait function in older adults.

RESEARCH PURPOSE

This study aimed to construct age-stratified normative data of non-dimensionally normalized temporal and spatial gait features in healthy older adults.

METHODS

We recruited 320 community-dwelling healthy adults aged 65 years or older from two prospective cohort studies. We stratified them into four age groups (65-69, 70-74, 75-79, and 80-84 years). Each age group comprised 40 men and 40 women. We obtained six gait features (cadence, step time, step time variability, step time asymmetry, gait speed, and step length) using a wearable inertia measurement unit attached on the skin overlying L3-L4 on the back. To mitigate the influence of body shape, we non-dimensionally normalized the gait features into unitless values using height and gravity.

RESULT

The effect of age group was significant in all raw gait features (p < 0.001 for step time variability, speed and step length; p < 0.05 for cadence, step time and step time asymmetry), and that of sex was significant in the five raw gait features, except for step time asymmetry(p < 0.001 for cadence, step time, speed, and step length; p < 0.05 for step time asymmetry). When gait features were normalized, the effect of age group remained (p < 0.001 for all gait features), whereas that of sex disappeared (p > 0.05 for all gait features).

SIGNIFICANCE

Our dimensionless normative data on gait features may be useful in comparative studies of gait function between sexes or ethnicities with different body shapes.

Development of a gait speed estimation model for healthy older adults using a single inertial measurement unit

Although gait speed changes are associated with various geriatric conditions, standard gait analysis systems, such as laboratory-based motion capture systems or instrumented walkways, are too expensive, spatially limited, and difficult to access. A wearable inertia sensor is cheap and easy to access; however, its accuracy in estimating gait speed is limited. In this study, we developed a model for accurately estimating the gait speed of healthy older adults using the data captured by an inertia sensor placed at their center of body mass (CoM). We enrolled 759 healthy older adults from two population-based cohort studies and asked them to walk on a 14 m long walkway thrice at comfortable paces with an inertia sensor attached to their CoM. In the middle of the walkway, we placed GAITRite™ to obtain the gold standard of gait speed. We then divided the participants into three subgroups using the normalized step length and developed a linear regression model for estimating the gold standard gait speed using age, foot length, and the features obtained from an inertia sensor, including cadence, vertical height displacement, yaw angle, and role angle of CoM. Our model exhibited excellent accuracy in estimating the gold standard gait speed (mean absolute error = 3.74%; root mean square error = 5.30 cm/s; intraclass correlation coefficient = 0.954). Our model may contribute to the early detection and monitoring of gait disorders and other geriatric conditions by making gait assessment easier, cheaper, and more ambulatory while remaining as accurate as other standard gait analysis systems.

Challenges in Quantifying Heel-Lift During Spacesuit Gait

INTRODUCTION: Heel-lift is a subjectively reported fit issue in planetary spacesuit boot prototypes that has not yet been quantified. Inertial measurement units (IMUs) could quantify heel-lift but are susceptible to integration drift. This work evaluates the use of IMUs and drift-correction algorithms, such as zero-velocity (ZVUs) and zero-position updates (ZPUs), to quantify heel-lift during spacesuited gait.METHODS: Data was originally collected by Fineman et al. in 2018 to assess lower body relative coordination in the spacesuit. IMUs were mounted on the spacesuit lower legs (SLLs) and spacesuit operator’s shank as three operators walked on a level walkway in three spacesuit padding conditions. Discrete wavelet transforms were used to identify foot-flat phase and heel-off for each step. Differences in heel-off timepoints were calculated in each step as a potential indicator of heel-lift, with spacesuit-delayed heel-off suggesting heel-lift. Average drift rates were estimated prior to and after applying ZVUs and ZPUs.RESULTS: Heel-off timepoint differences showed instances of spacesuit-delayed heel-off and instances of operator-delayed heel-off. Drift rates after applying ZVUs and ZPUs suggested an upper time bound of 0.03 s past heel-off to measure heel-lift magnitude with an accuracy of 1 cm.DISCUSSION: Results suggest that IMUs may not be appropriate for quantifying the presence and magnitude of heel lift. Operator-delayed heel-off suggests that the SLL may be expanding prior to heel-off, creating a false vertical acceleration signal interpreted by this study to be spacesuit heel-off. Quantifying heel-off will therefore require improvements in IMU mounting to mitigate the effects of SLL, or alternative sensor technologies.Boppana A, Priddy ST, Stirling L, Anderson AP. Challenges in quantifying heel-lift during spacesuit gait. Aerosp Med Hum Perform. 2022; 93(8):643–648.

Using Sensor Graphs for Monitoring the Effect on the Performance of the OTAGO Exercise Program in Older Adults

The OTAGO exercise program is effective in decreasing the risk for falls of older adults. This research investigated if there is an indication that the OTAGO exercise program has a positive effect on the capacity and as well as on the performance in mobility. We used the data of the 10-months observational OTAGO pilot study with 15 (m = 1, f = 14) (pre-)frail participants aged 84.60 y (SD: 5.57 y). Motion sensors were installed in the flats of the participants and used to monitor their activity as a surrogate variable for performance. We derived a weighted directed multigraph from the physical sensor network, subtracted the weights of one day from a baseline, and used the difference in percent to quantify the change in performance. Least squares was used to compute the overall progress of the intervention (n = 9) and the control group (n = 6). In accordance with previous studies, we found indication for a positive effect of the OTAGO program on the capacity in both groups. Moreover, we found indication that the OTAGO program reduces the decline in performance of older adults in daily living. However, it is too early to conclude causalities from our findings because the data was collected during a pilot study.

A Pilot Study to Validate a Wearable Inertial Sensor for Gait Assessment in Older Adults with Falls

The high prevalence of falls and the enormous impact they have on the elderly population is a cause for concern. We aimed to develop a walking-monitor gait pattern (G-STRIDE) for older adults based on a 6-axis inertial measurement (IMU) with the application of pedestrian dead reckoning algorithms and tested its structural and clinical validity. A cross-sectional case–control study was conducted with 21 participants (11 fallers and 10 non-fallers). We measured gait using an IMU attached to the foot while participants walked around different grounds (indoor flooring, outdoor floor, asphalt, etc.). The G-STRIDE consisted of a portable inertial device that monitored the gait pattern and a mobile app for telematic clinical analysis. G-STRIDE made it possible to measure gait parameters under normal living conditions when walking without assessing the patient in the outpatient clinic. Moreover, we verified concurrent validity with convectional outcome measures using intraclass correlation coefficients (ICCs) and analyzed the differences between participants. G-STRIDE showed high estimation accuracy for the walking speed of the elderly and good concurrent validity compared to conventional measures (ICC = 0.69; p < 0.000). In conclusion, the developed inertial-based G-STRIDE can accurately classify older people with risk to fall with a significance as high as using traditional but more subjective clinical methods (gait speed, Timed Up and Go Test).

Orientation-Invariant Spatio-Temporal Gait Analysis Using Foot-Worn Inertial Sensors

Inertial sensors can potentially assist clinical decision making in gait-related disorders. Methods for objective spatio-temporal gait analysis usually assume the careful alignment of the sensors on the body, so that sensor data can be evaluated using the body coordinate system. Some studies infer sensor orientation by exploring the cyclic characteristics of walking. In addition to being unrealistic to assume that the sensor can be aligned perfectly with the body, the robustness of gait analysis with respect to differences in sensor orientation has not yet been investigated-potentially hindering use in clinical settings. To address this gap in the literature, we introduce an orientation-invariant gait analysis approach and propose a method to quantitatively assess robustness to changes in sensor orientation. We validate our results in a group of young adults, using an optical motion capture system as reference. Overall, good agreement between systems is achieved considering an extensive set of gait metrics. Gait speed is evaluated with a relative error of -3.1±9.2 cm/s, but precision improves when turning strides are excluded from the analysis, resulting in a relative error of -3.4±6.9 cm/s. We demonstrate the invariance of our approach by simulating rotations of the sensor on the foot.

Toward a remote assessment of walking bout and speed: application in patients with multiple sclerosis

Gait speed as a powerful biomarker of mobility is mostly assessed in controlled environments, e.g. in the clinic. With wearable inertial sensors, gait speed can be estimated in an objective manner. However, most of the previous works have validated the gait speed estimation algorithms in clinical settings which can be different than the home assessments in which the patients demonstrate their actual performance. Moreover, to provide comfort for the users, devising an algorithm based on a single sensor setup is essential. To this end, the goal of this study was to develop and validate a new gait speed estimation method based on a machine learning approach to predict gait speed in both clinical and home assessments by a sensor on the lower back. Moreover, two methods were introduced to detect walking bouts during daily activities at home. We have validated the algorithms in 35 patients with multiple sclerosis as it often presents with mobility difficulties. Therefore, the robustness of the algorithm can be shown in an impaired or slow gait. Against silver standard multi-sensor references, we achieved a bias close to zero and a precision of 0.15 m/s for gait speed estimation. Furthermore, the proposed machine learning-based locomotion detection method had a median of 96.8% specificity, 93.0% sensitivity, 96.4% accuracy, and 78.6% F1-score in detecting walking bouts at home. The high performance of the proposed algorithm showed the feasibility of the unsupervised mobility assessment introduced in this study.

Validation of Walking Speed Estimation from Trunk Mounted Accelerometers for a Range of Walking Speeds

Walking speed is a strong indicator of the health status of older people and patients. Using algorithms, the walking speed can be estimated from wearable accelerometers, which enables minimally obtrusive (longitudinal) monitoring. We evaluated the performance of two algorithms, the inverted pendulum (IP) algorithm, and a novel adaptation correcting for lateral step movement, which aimed to improve accuracy during slow walking. To evaluate robustness, we gathered data from different groups (healthy adults, elderly, and elderly patients) of volunteers (n = 159) walking under various conditions (over ground, treadmill, using walking aids) at a broad range of speeds (0.11–1.93 m/s). Both of the algorithms showed good agreement with the reference values and similar root-mean-square errors (RMSEs) for walking speeds ≥0.5 m/s, which ranged from 0.09–0.16 m/s for the different positions, in line with the results from others. However, for slower walking, RMSEs were significantly better for the new method (0.06–0.09 m/s versus 0.15–0.19 m/s). Pearson correlation improved for speeds <0.5 m/s (from 0.67–0.72 to 0.73–0.82) as well as higher speeds (0.87–0.97 to 0.90–0.98) with the new method. Overall, we found that IP(-based) walking speed estimation proved to be applicable for a variety of wearing positions, conditions and speeds, indicating its potential value for health assessment applications.

A Deep Learning Approach for TUG and SPPB Score Prediction of (Pre-) Frail Older Adults on Real-Life IMU Data

Since older adults are prone to functional decline, using Inertial-Measurement-Units (IMU) for mobility assessment score prediction gives valuable information to physicians to diagnose changes in mobility and physical performance at an early stage and increases the chances of rehabilitation. This research introduces an approach for predicting the score of the Timed Up & Go test and Short-Physical-Performance-Battery assessment using IMU data and deep neural networks. The approach is validated on real-world data of a cohort of 20 frail or (pre-) frail older adults of an average of 84.7 years. The deep neural networks achieve an accuracy of about 95% for both tests for participants known by the network.

Using Different Combinations of Body-Mounted IMU Sensors to Estimate Speed of Horses-A Machine Learning Approach

Speed is an essential parameter in biomechanical analysis and general locomotion research. It is possible to estimate the speed using global positioning systems (GPS) or inertial measurement units (IMUs). However, GPS requires a consistent signal connection to satellites, and errors accumulate during IMU signals integration. In an attempt to overcome these issues, we have investigated the possibility of estimating the horse speed by developing machine learning (ML) models using the signals from seven body-mounted IMUs. Since motion patterns extracted from IMU signals are different between breeds and gaits, we trained the models based on data from 40 Icelandic and Franches-Montagnes horses during walk, trot, tölt, pace, and canter. In addition, we studied the estimation accuracy between IMU locations on the body (sacrum, withers, head, and limbs). The models were evaluated per gait and were compared between ML algorithms and IMU location. The model yielded the highest estimation accuracy of speed (RMSE = 0.25 m/s) within equine and most of human speed estimation literature. In conclusion, highly accurate horse speed estimation models, independent of IMU(s) location on-body and gait, were developed using ML.