Use of a Single Wireless IMU for the Segmentation and Automatic Analysis of Activities Performed in the 3-m Timed Up & Go Test

Implementing and Evaluating the Timed Up and Go Test Automation Using Smartphones and Smartwatches

Physical performance tests aim to assess the physical abilities and mobility skills of individuals for various healthcare purposes. They are often driven by experts and usually performed at their practice, and therefore they are resource-intensive and time-demanding. For tests based on objective measurements (e.g., duration, repetitions), technology can be used to automate them, allowing the patients to perform the test themselves, more frequently and anywhere, while alleviating the expert from supervising the test. The well-known Timed Up and Go (TUG) test, typically used for mobility assessment, is an ideal candidate for automation, as inertial sensors (among others) can be deployed to detect the various movements constituting the test without expert supervision. To move from expert-led testing to self-administered testing, we present a mHealth system capable of automating the TUG test using a pocket-sized smartphone or a wrist smartwatch paired with a smartphone, where data from inertial sensors are used to detect the activities carried out by the patient while performing the test and compute their results in real time. All processing (i.e., data processing, machine learning-based activity inference, results calculation) takes place on the smartphone. The use of both devices to automate the TUG test was evaluated (w.r.t. accuracy, reliability and battery consumption) and mutually compared, and set off with a reference method, obtaining excellent Bland-Altman agreement results and Intraclass Correlation Coefficient reliability. Results also suggest that the smartwatch-based system performs better than the smartphone-based system.

L Test Subtask Segmentation for Lower-Limb Amputees Using a Random Forest Algorithm

Functional mobility tests, such as the L test of functional mobility, are recommended to provide clinicians with information regarding the mobility progress of lower-limb amputees. Smartphone inertial sensors have been used to perform subtask segmentation on functional mobility tests, providing further clinically useful measures such as fall risk. However, L test subtask segmentation rule-based algorithms developed for able-bodied individuals have not produced sufficiently acceptable results when tested with lower-limb amputee data. In this paper, a random forest machine learning model was trained to segment subtasks of the L test for application to lower-limb amputees. The model was trained with 105 trials completed by able-bodied participants and 25 trials completed by lower-limb amputee participants and tested using a leave-one-out method with lower-limb amputees. This algorithm successfully classified subtasks within a one-foot strike for most lower-limb amputee participants. The algorithm produced acceptable results to enhance clinician understanding of a person's mobility status (>85% accuracy, >75% sensitivity, >95% specificity).

Investigating the Effects of Virtual Reality-Based Training on Balance Ability and Balance Confidence in Older Individuals

Each year, over 25% of adults aged sixty-five years old or older suffer a fall, and three million are treated for fall-related injuries due to lack of balance. Here, we aimed to investigate how virtual reality (VR)-based training affects balance performance and confidence in older adults. To accomplish this goal, we studied 21 healthy, older individuals between 60 and 85 years old, both pre- and post-training (6 weeks of training, twice per week (or 12 sessions)). The VR group donned an Oculus VR headset and consisted of nine participants (aged 75.9 ± 3.7 years old), while the control group (aged 75.1 ± 6.7 years old) performed training without a headset and consisted of eight participants that completed our study. To assess balance ability, we utilized the Balance Error Scoring System (BESS) and the Timed Up and Go (TUG) test. To assess balance confidence, we implemented the Activities-Specific Balance Confidence (ABC) Scale and, to assess fear of falling, the Tinetti Falls Efficacy Scale (FES). Further, we assessed depression (via the Geriatric Depression Scale (GDS)) and cognitive ability (via the Mini-Mental State Examination (MMSE)). The post-training results showed improvements in balance ability for both the VR and control groups, as well as changes in the relationship between balance confidence and balance ability for the VR group only. Further, improvements in cognitive ability were seen in the control group. This study is an indication that older individuals' balance ability may benefit from several weeks of targeted training.

Quantifying Turning Tasks With Wearable Sensors: A Reliability Assessment

OBJECTIVE

The aim of this study was to establish the test-retest reliability of metrics obtained from wearable inertial sensors that reflect turning performance during tasks designed to imitate various turns in daily activity.

METHODS

Seventy-one adults who were healthy completed 3 turning tasks: a 1-minute walk along a 6-m walkway, a modified Illinois Agility Test (mIAT), and a complex turning course (CTC). Peak axial turning and rotational velocity (yaw angular velocity) were extracted from wearable inertial sensors on the head, trunk, and lumbar spine. Intraclass correlation coefficients (ICCs) were established to assess the test-retest reliability of average peak turning speed for each task. Lap time was collected for reliability analysis as well.

RESULTS

Turning speed across all tasks demonstrated good to excellent reliability, with the highest reliability noted for the CTC (45-degree turns: ICC = 0.73-0.81; 90-degree turns: ICC = 0.71-0.83; and 135-degree turns: ICC = 0.72-0.80). The reliability of turning speed during 180-degree turns from the 1-minute walk was consistent across all body segments (ICC = 0.74-0.76). mIAT reliability ranged from fair to excellent (end turns: ICC = 0.52-0.72; mid turns: ICC = 0.50-0.56; and slalom turns: ICC = 0.66-0.84). The CTC average lap time demonstrated good test-retest reliability (ICC = 0.69), and the mIAT average lap time test-retest reliability was excellent (ICC = 0.91).

CONCLUSION

Turning speed measured by inertial sensors is a reliable outcome across a variety of ecologically valid turning tasks that can be easily tested in a clinical environment.

IMPACT

Turning performance is a reliable and important measure that should be included in clinical assessments and clinical trials.

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.

Subtask Segmentation Methods of the Timed Up and Go Test and L Test Using Inertial Measurement Units—A Scoping Review

The Timed Up and Go test (TUG) and L Test are functional mobility tests that allow healthcare providers to assess a person’s balance and fall risk. Segmenting these mobility tests into their respective subtasks, using sensors, can provide further and more precise information on mobility status. To identify and compare current methods for subtask segmentation using inertial sensor data, a scoping review of the literature was conducted using PubMed, Scopus, and Google Scholar. Articles were identified that described subtask segmentation methods for the TUG and L Test using only inertial sensor data. The filtering method, ground truth estimation device, demographic, and algorithm type were compared. One article segmenting the L Test and 24 articles segmenting the TUG met the criteria. The articles were published between 2008 and 2022. Five studies used a mobile smart device’s inertial measurement system, while 20 studies used a varying number of external inertial measurement units. Healthy adults, people with Parkinson’s Disease, and the elderly were the most common demographics. A universally accepted method for segmenting the TUG test and the L Test has yet to be published. Angular velocity in the vertical and mediolateral directions were common signals for subtask differentiation. Increasing sample sizes and furthering the comparison of segmentation methods with the same test sets will allow us to expand the knowledge generated from these clinically accessible tests.

Deep Learning-Based Subtask Segmentation of Timed Up-and-Go Test Using RGB-D Cameras

The timed up-and-go (TUG) test is an efficient way to evaluate an individual's basic functional mobility, such as standing up, walking, turning around, and sitting back. The total completion time of the TUG test is a metric indicating an individual's overall mobility. Moreover, the fine-grained consumption time of the individual subtasks in the TUG test may provide important clinical information, such as elapsed time and speed of each TUG subtask, which may not only assist professionals in clinical interventions but also distinguish the functional recovery of patients. To perform more accurate, efficient, robust, and objective tests, this paper proposes a novel deep learning-based subtask segmentation of the TUG test using a dilated temporal convolutional network with a single RGB-D camera. Evaluation with three different subject groups (healthy young, healthy adult, stroke patients) showed that the proposed method demonstrated better generality and achieved a significantly higher and more robust performance (healthy young = 95.458%, healthy adult = 94.525%, stroke = 93.578%) than the existing rule-based and artificial neural network-based subtask segmentation methods. Additionally, the results indicated that the input from the pelvis alone achieved the best accuracy among many other single inputs or combinations of inputs, which allows a real-time inference (approximately 15 Hz) in edge devices, such as smartphones.

A Novel Privacy Preservation and Quantification Methodology for Implementing Home-Care-Oriented Movement Analysis Systems

Human movement is generally evaluated through both observations and clinical assessment scales to identify the state and deterioration of a patient's motor control. Lately, technological systems for human motion analysis have been used in clinics to identify abnormal movement states, while they generally suffer from privacy challenges and concerns especially at home or in remote places. This paper presents a novel privacy preservation and quantification methodology that imitates the forgetting process of human memory to protect privacy in patient-centric healthcare. The privacy preservation principle of this methodology is to change the traditional data analytic routines into a distributed and disposable form (i.e., DnD) so as to naturally minimise the disclosure of patients' health data. To help judge the efficacy of DnD-based privacy preservation, the researchers further developed a risk-driven privacy quantification framework to supplement the existing privacy quantification techniques. To facilitate validating the methodology, this research also involves a home-care-oriented movement analysis system that comprises a single inertial measurement sensor and a mobile application. The system can acquire personal information, raw data of movements and indexes to evaluate the risk of falls and gait at homes. Moreover, the researchers conducted a technological appreciation survey of 16 health professionals to help understand the perception of this research. The survey obtains positive feedback regarding the movement analysis system and the proposed methodology as suitable for home-care scenarios.

Machine learning classification of multiple sclerosis patients based on raw data from an instrumented walkway

Background

Using embedded sensors, instrumented walkways provide clinicians with important information regarding gait disturbances. However, because raw data are summarized into standard gait variables, there may be some salient features and patterns that are ignored. Multiple sclerosis (MS) is an inflammatory neurodegenerative disease which predominantly impacts young to middle-aged adults. People with MS may experience varying degrees of gait impairments, making it a reasonable model to test contemporary machine leaning algorithms. In this study, we employ machine learning techniques applied to raw walkway data to discern MS patients from healthy controls. We achieve this goal by constructing a range of new features which supplement standard parameters to improve machine learning model performance.

Results

Eleven variables from the standard gait feature set achieved the highest accuracy of 81%, precision of 95%, recall of 81%, and F1-score of 87%, using support vector machine (SVM). The inclusion of the novel features (toe direction, hull area, base of support area, foot length, foot width and foot area) increased classification accuracy by 7%, recall by 9%, and F1-score by 6%.

Conclusions

The use of an instrumented walkway can generate rich data that is generally unseen by clinicians and researchers. Machine learning applied to standard gait variables can discern MS patients from healthy controls with excellent accuracy. Noteworthy, classifications are made stronger by including novel gait features (toe direction, hull area, base of support area, foot length and foot area).

Characterization of stroke-related upper limb motor impairments across various upper limb activities by use of kinematic core set measures

Background

Upper limb kinematic assessments provide quantifiable information on qualitative movement behavior and limitations after stroke. A comprehensive characterization of spatiotemporal kinematics of stroke subjects during upper limb daily living activities is lacking. Herein, kinematic expressions were investigated with respect to different movement types and impairment levels for the entire task as well as for motion subphases.

Method

Chronic stroke subjects with upper limb movement impairments and healthy subjects performed a set of daily living activities including gesture and grasp movements. Kinematic measures of trunk displacement, shoulder flexion/extension, shoulder abduction/adduction, elbow flexion/extension, forearm pronation/supination, wrist flexion/extension, movement time, hand peak velocity, number of velocity peaks (NVP), and spectral arc length (SPARC) were extracted for the whole movement as well as the subphases of reaching distally and proximally. The effects of the factors gesture versus grasp movements, and the impairment level on the kinematics of the whole task were tested. Similarities considering the metrics expressions and relations were investigated for the subphases of reaching proximally and distally between tasks and subgroups.

Results

Data of 26 stroke and 5 healthy subjects were included. Gesture and grasp movements were differently expressed across subjects. Gestures were performed with larger shoulder motions besides higher peak velocity. Grasp movements were expressed by larger trunk, forearm, and wrist motions. Trunk displacement, movement time, and NVP increased and shoulder flexion/extension decreased significantly with increased impairment level. Across tasks, phases of reaching distally were comparable in terms of trunk displacement, shoulder motions and peak velocity, while reaching proximally showed comparable expressions in trunk motions. Consistent metric relations during reaching distally were found between shoulder flexion/extension, elbow flexion/extension, peak velocity, and between movement time, NVP, and SPARC. Reaching proximally revealed reproducible correlations between forearm pronation/supination and wrist flexion/extension, movement time and NVP.

Conclusion

Spatiotemporal differences between gestures versus grasp movements and between different impairment levels were confirmed. The consistencies of metric expressions during movement subphases across tasks can be useful for linking kinematic assessment standards and daily living measures in future research and performing task and study comparisons.

Trial registration: ClinicalTrials.gov Identifier NCT03135093. Registered 26 April 2017, https://clinicaltrials.gov/ct2/show/NCT03135093.

Instrumental Timed Up and Go test discloses abnormalities in patients with Cervical Dystonia

Background Cervical dystonia is a movement disorder characterized by involuntary and sustained contraction of the neck muscles that determines abnormal posture. The aim of this study was to investigate whether dystonic posture in patients with cervical dystonia affects walking and causes postural changes. Methods Patients with cervical dystonia and a group of age-matched healthy controls underwent an instrumental evaluation of the Timed Up and Go Test. Findings All the spatio-temporal parameters of the sub-phases of the Timed up and go test had a significantly higher duration in cervical dystonia patients compared to the control group while no differences in flection and extension angular amplitudes were observed. Indeed, we found that Cervical Dystonia patients had abnormalities in turning, as well as in standing-up and sitting-down from a chair during the Timed up and go test than healthy controls. Interpretation Impairment in postural control in cervical dystonia patients during walking and postural changes prompts to develop rehabilitation strategies to improve postural stability and reduce the risk of fall in these patients.

Between-site equivalence of turning speed assessments using inertial measurement units

BACKGROUND

Turning is a component of gait that requires planning for movement of multiple body segments and the sophisticated integration of sensory information from the vestibular, visual, and somatosensory systems. These aspects of turning have led to growing interest to quantify turning in clinical populations to characterize deficits or identify disease progression. However, turning may be affected by environmental differences, and the degree to which turning assessments are comparable across research or clinical sites has not yet been evaluated.

RESEARCH QUESTION

The aim of this study was to determine the extent to which peak turning speeds are equivalent between two sites for a variety of mobility tasks.

METHODS

Data were collected at two different sites using separate healthy young adult participants (n = 47 participants total), but recruited using identical inclusion and exclusion criteria. Participants at each site completed three turning tasks: a one-minute walk (1 MW) along a six-meter walkway, a modified Illinois Agility Test (mIAT), and a custom clinical turning course (CCTC). Peak yaw turning speeds were extracted from wearable inertial sensors on the head, trunk, and pelvis. Between-site differences and two one-sided tests (TOST) were used to determine equivalence between sites, based on a minimum effect size reported between individuals with mild traumatic brain injury and healthy control subjects.

RESULTS

No outcomes were different between sites, and equivalence was determined for 6/21 of the outcomes. These findings suggest that some turning tasks and outcome measures may be better suited for multi-site studies. The equivalence results are also dependent on the minimum effect size of interest; nearly all outcomes were equivalent across sites when larger minimum effect sizes of interest were used.

SIGNIFICANCE

Together, these results suggest some tasks and outcome measures may be better suited for multi-site studies and literature-based comparisons.

Automatic Functional Shoulder Task Identification and Sub-task Segmentation Using Wearable Inertial Measurement Units for Frozen Shoulder Assessment

Advanced sensor technologies have been applied to support frozen shoulder assessment. Sensor-based assessment tools provide objective, continuous and quantitative information for evaluation and diagnosis. However, the current tools for assessment of functional shoulder tasks mainly rely on manual operation. It may cause several technical issues to the reliability and usability of the assessment tool, including manual bias during the recording and additional efforts for data labeling. To tackle these issues, this pilot study aims to propose an automatic functional shoulder task identification and sub-task segmentation system using inertial measurement units to provide reliable shoulder task labeling and sub-task information for clinical professionals. The proposed method combines machine learning models and rule-based modification to identify shoulder tasks and segment sub-tasks accurately. A hierarchical design is applied to enhance the efficiency and performance of the proposed approach. Nine healthy subjects and nine frozen shoulder patients are invited to perform five common shoulder tasks in the lab-based and clinical environments, respectively. The experimental results show that the proposed method can achieve 87.11% F-score for shoulder task identification, and 83.23% F-score and 427 mean absolute time errors (milliseconds) for sub-task segmentation. The proposed approach demonstrates the feasibility of the proposed method to support reliable evaluation for clinical assessment.

A Novel Capacitive Step Sensor to Trigger Stimulation on Functional Electrical Stimulators Devices for Drop Foot

Drop foot is a typical clinical condition associated with stroke. According to the World Health Organization, fifteen million people suffer a stroke per year, and one of three people's survival gets drop foot. Functional Electrical Stimulation systems are applied over the peroneal motor nerve to achieve the drop foot problem's dorsiflexion. An accurate and reliable way to identify in real-time the gait phases to trigger and finish the stimulation is needed. This paper proposes a new step sensor with a custom capacitive pressure sensors array located under the heel to detect a gait pattern in real-time to synchronize the stimulation with the user gait. The step sensor uses a capacitive pressure sensors array and hardware, which acquire the signals, execute an algorithm to detect the start and finish of the swing phase in real-time, and send the synchronization signal wirelessly. The step sensor was tested in two ways: 10 meters walk test and walking in a treadmill for 2 minutes. These two tests were performed with two different walk velocities and with thirteen healthy volunteers. Thus, all the 1342 steps were correctly detected. Compared to an inertial sensor located in the lower-back, the proposed step sensor achieves a mean error of 27.60±0.03 [ms] for the detection of the start of the swing phase and a mean error of 20.86±0.02 [ms] for the detection of the end of the swing phase. The results show an improvement in time error (respect to others pressure step sensors), sensibility and specificity (both 100%), and comfortability.

Inertial Sensor Algorithms to Characterize Turning in Neurological Patients With Turn Hesitations

BACKGROUND

One difficulty in turning algorithm design for inertial sensors is detecting two discrete turns in the same direction, close in time. A second difficulty is under-estimation of turn angle due to short-duration hesitations by people with neurological disorders. We aimed to validate and determine the generalizability of a: I. Discrete Turn Algorithm for variable and sequential turns close in time and II: Merged Turn Algorithm for a single turn angle in the presence of hesitations.

METHODS

We validated the Discrete Turn Algorithm with motion capture in healthy controls (HC, n = 10) performing a spectrum of turn angles. Subsequently, the generalizability of the Discrete Turn Algorithm and associated, Merged Turn Algorithm were tested in people with Parkinson's disease (PD, n = 124), spinocerebellar ataxia (SCA, n = 51), and HC (n = 125).

RESULTS

The Discrete Turn Algorithm shows improved agreement with optical motion capture and with known turn angles, compared to our previous algorithm by El-Gohary et al. The Merged Turn algorithm that merges consecutive turns in the same direction with short hesitations resulted in turn angle estimates closer to a fixed 180-degree turn angle in the PD, SCA, and HC subjects compared to our previous turn algorithm. Additional metrics were proposed to capture turn hesitations in PD and SCA.

CONCLUSION

The Discrete Turn Algorithm may be particularly useful to characterize turns when the turn angle is unknown, i.e., during free-living conditions. The Merged Turn algorithm is recommended for clinical tasks in which the single-turn angle is known, especially for patients who hesitate while turning.

Construct validity and test-retest reliability of a free mobile application for spatio-temporal gait analysis in Parkinson's disease patients

BACKGROUND

Mobile applications may be used to assess gait pattern deviation through mobile smartphones in people with Parkinson's disease (PD). However, few studies have investigated their psychometrics properties.

RESEARCH QUESTION

To study the construct validity and test-retest reliability of the RUNZI® free mobile application in people with mild to moderate PD.

METHODS

Thirty individuals were evaluated with the RUNZI® app and with the 10-meter walking test (10 MW), simultaneously. In addition, the Timed Up & Go test (TUG), Tinetti scale, and the Berg Balance Scale (BBS) were used to study the construct validity. Also, test-retest reliability of the mobile for spatio-temporal gait parameters was explored.

RESULTS

The correlation indices of the 10 MW test with the RUNZI® app at fast speeds was moderate to excellent (r = .588-.957). At a comfortable speed, the correlation was excellent for walking speed (r = 0.944), moderate for steps (r = 0.780) and stride length (r = 0.760), and poor for cadence (r = .424). Results showed significant correlations between TUG and spatio-temporal gait parameters at fast and comfortable speeds. There were no significant correlations or consistent associations between Tinetti and BBS and RUNZI®. The test-retest reliability was good to excellent for parameters measured with the RUNZI®.

SIGNIFICANCE

Our findings highlight specific opportunities for a free smartphone-based spatio-temporal gait analysis to serve as a complement to conventional gait analysis methods in clinical practice with a moderate to excellent construct validity with the 10 MW test and good to excellent test-retest reliability in PD patients.