Timed Up-and-Go phase segmentation in Parkinson's disease patients using unobtrusive inertial sensors

Wearable sensor-based gait analysis to discriminate early Parkinson's disease from essential tremor

BACKGROUND

Differentiating early-stage Parkinson's disease (PD) from essential tremor (ET) is challenging since they have some overlapping clinical features. Since early-stage PD may present with slight gait impairment and ET generally does not, gait analysis could be used to differentiate PD from ET using machine learning.

OBJECTIVE

To differentiate early-stage PD from ET via machine learning using gait and postural transition parameters calculated using the raw kinematic signal captured from inertial measurement unit (IMU) sensors.

METHODS

Gait and postural transition parameters were collected from 84 early-stage PD and 80 ET subjects during the Time Up and Go (TUG) test. We randomly split our data into training and test data. Within the training data, we separated the TUG test into four components: standing, straight walk, turning, and sitting to build weighted average ensemble classification models. The four components' weight indices were trained using logistic regression. Several ensemble models' leave-one-out cross-validation (LOOCV) performances were compared. Independent test data were used to evaluate the model with the best LOOCV performance.

RESULTS

The best weighted average ensemble classification model LOOCV results included an accuracy of 84%, Kappa of 0.68, sensitivity of 85.9%, specificity of 82.1%, and AUC of 0.912. Thirty-three gait and postural transition parameters, such as Arm-Symbolic Symmetry Index and 180° Turn-Max Angular Velocity, were included in Feature Group III. The independent test data achieved a 75.8% accuracy.

CONCLUSIONS

Our findings suggest that gait and postural transition parameters obtained from wearable sensors combined with machine learning had the potential to distinguish between early-stage PD and ET.

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.

Wearable gait analysis systems: ready to be used by medical practitioners in geriatric wards?

Aim

To investigate the feasibility of wearable gait analysis in geriatric wards by testing the effectiveness and acceptance of the system.

Findings

Wearable gait analysis can be implemented into geriatric wards, showing its readiness for a transformation from a pure research tool to a practically usable gait analysis system.

Message

Despite good transferability into clinical practice, future research should aim to increase functionality and applicability of wearable gait analysis systems in clinical contexts.

Purpose

We assess feasibility of wearable gait analysis in geriatric wards by testing the effectiveness and acceptance of the system.

Methods

Gait parameters of 83 patients (83.34 ± 5.88 years, 58/25 female/male) were recorded at admission and/or discharge to/from two geriatric inpatient wards. Gait parameters were tested for statistically significant differences between admission and discharge. Walking distance measured by a wearable gait analysis system was correlated with distance assessed by physiotherapists. Examiners rated usability using the system usability scale. Patients reported acceptability on a five-point Likert-scale.

Results

The total distance measures highly correlate (r = 0.89). System Usability Scale is above the median threshold of 68, indicating good usability. Majority of patients does not have objections regarding the use of the system. Among other gait parameters, mean heel strike angle changes significantly between admission and discharge.

Conclusion

Wearable gait analysis system is objectively and subjectively usable in a clinical setting and accepted by patients. It offers a reasonably valid assessment of gait parameters and is a feasible way for instrumented gait analysis.

Reliability of the infrared motion-time acquisition system for each motion segment in the timed up-and-go test

[Purpose] This study aimed to investigate the reliability of an infrared motion-time acquisition system by measuring the time taken for five motion segments (sit-to-stand, forward gait, mid-turn, return gait, and turn-stand-to-sit) in the timed up-and-go test. [Participants and Methods] In total, 30 healthy adults (25.1 ± 4.6 years, 19 males and 11 females) were included in this study. Tester A and Tester B measured the time taken in the timed up-and-go test and its five motion segments with an infrared motion-time acquisition system, and two measurements were made by Tester A and one by Tester B. [Results] Intraclass correlation coefficients of the time taken for the five motion segments in the timed up-and-go test and the intra- and inter-rater intraclass correlation coefficients were greater than 0.9. [Conclusion] Infrared motion-time acquisition systems and its five motion segments are reliable and provide accurate measurements during the timed up-and-go test.

A deep explainable artificial intelligent framework for neurological disorders discrimination

Pathological hand tremor (PHT) is a common symptom of Parkinson's disease (PD) and essential tremor (ET), which affects manual targeting, motor coordination, and movement kinetics. Effective treatment and management of the symptoms relies on the correct and in-time diagnosis of the affected individuals, where the characteristics of PHT serve as an imperative metric for this purpose. Due to the overlapping features of the corresponding symptoms, however, a high level of expertise and specialized diagnostic methodologies are required to correctly distinguish PD from ET. In this work, we propose the data-driven [Formula: see text] model, which processes the kinematics of the hand in the affected individuals and classifies the patients into PD or ET. [Formula: see text] is trained over 90 hours of hand motion signals consisting of 250 tremor assessments from 81 patients, recorded at the London Movement Disorders Centre, ON, Canada. The [Formula: see text] outperforms its state-of-the-art counterparts achieving exceptional differential diagnosis accuracy of [Formula: see text]. In addition, using the explainability and interpretability measures for machine learning models, clinically viable and statistically significant insights on how the data-driven model discriminates between the two groups of patients are achieved.

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.

Wearable Health Technology to Quantify the Functional Impact of Peripheral Neuropathy on Mobility in Parkinson's Disease: A Systematic Review

The occurrence of peripheral neuropathy (PNP) is often observed in Parkinson’s disease (PD) patients with a prevalence up to 55%, leading to more prominent functional deficits. Motor assessment with mobile health technologies allows high sensitivity and accuracy and is widely adopted in PD, but scarcely used for PNP assessments. This review provides a comprehensive overview of the methodologies and the most relevant features to investigate PNP and PD motor deficits with wearables. Because of the lack of studies investigating motor impairments in this specific subset of PNP-PD patients, Pubmed, Scopus, and Web of Science electronic databases were used to summarize the state of the art on PNP motor assessment with wearable technology and compare it with the existing evidence on PD. A total of 24 papers on PNP and 13 on PD were selected for data extraction: The main characteristics were described, highlighting major findings, clinical applications, and the most relevant features. The information from both groups (PNP and PD) was merged for defining future directions for the assessment of PNP-PD patients with wearable technology. We established suggestions on the assessment protocol aiming at accurate patient monitoring, targeting personalized treatments and strategies to prevent falls and to investigate PD and PNP motor characteristics.

Subtask Segmentation of Timed Up and Go Test for Mobility Assessment of Perioperative Total Knee Arthroplasty

Total knee arthroplasty (TKA) is one of the most common treatments for people with severe knee osteoarthritis (OA). The accuracy of outcome measurements and quantitative assessments for perioperative TKA is an important issue in clinical practice. Timed up and go (TUG) tests have been validated to measure basic mobility and balance capabilities. A TUG test contains a series of subtasks, including sit-to-stand, walking-out, turning, walking-in, turning around, and stand-to-sit tasks. Detailed information about subtasks is essential to aid clinical professionals and physiotherapists in making assessment decisions. The main objective of this study is to design and develop a subtask segmentation approach using machine-learning models and knowledge-based postprocessing during the TUG test for perioperative TKA. The experiment recruited 26 patients with severe knee OA (11 patients with bilateral TKA planned and 15 patients with unilateral TKA planned). A series of signal-processing mechanisms and pattern recognition approaches involving machine learning-based multi-classifiers, fragmentation modification and subtask inference are designed and developed to tackle technical challenges in typical classification algorithms, including motion variability, fragmentation and ambiguity. The experimental results reveal that the accuracy of the proposed subtask segmentation approach using the AdaBoost technique with a window size of 128 samples is 92%, which is an improvement of at least 15% compared to that of the typical subtask segmentation approach using machine-learning models only.

Optimization of IMU Sensor Placement for the Measurement of Lower Limb Joint Kinematics

There is an increased interest in using wearable inertial measurement units (IMUs) in clinical contexts for the diagnosis and rehabilitation of gait pathologies. Despite this interest, there is a lack of research regarding optimal sensor placement when measuring joint kinematics and few studies which examine functionally relevant motions other than straight level walking. The goal of this clinical measurement research study was to investigate how the location of IMU sensors on the lower body impact the accuracy of IMU-based hip, knee, and ankle angular kinematics. IMUs were placed on 11 different locations on the body to measure lower limb joint angles in seven participants performing the timed-up-and-go (TUG) test. Angles were determined using different combinations of IMUs and the TUG was segmented into different functional movements. Mean bias and root mean square error values were computed using generalized estimating equations comparing IMU-derived angles to a reference optical motion capture system. Bias and RMSE values vary with the sensor position. This effect is partially dependent on the functional movement analyzed and the joint angle measured. However, certain combinations of sensors produce lower bias and RMSE more often than others. The data presented here can inform clinicians and researchers of placement of IMUs on the body that will produce lower error when measuring joint kinematics for multiple functionally relevant motions. Optimization of IMU-based kinematic measurements is important because of increased interest in the use of IMUs to inform diagnose and rehabilitation in clinical settings and at home.

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

Falls represent a major public health problem in the elderly population. The Timed Up & Go test (TU & Go) is the most used tool to measure this risk of falling, which offers a unique parameter in seconds that represents the dynamic balance. However, it is not determined in which activity the subject presents greater difficulties. For this, a feature-based segmentation method using a single wireless Inertial Measurement Unit (IMU) is proposed in order to analyze data of the inertial sensors to provide a complete report on risks of falls. Twenty-five young subjects and 12 older adults were measured to validate the method proposed with an IMU in the back and with video recording. The measurement system showed similar data compared to the conventional test video recorded, with a Pearson correlation coefficient of 0.9884 and a mean error of 0.17 ± 0.13 s for young subjects, as well as a correlation coefficient of 0.9878 and a mean error of 0.2 ± 0.22 s for older adults. Our methodology allows for identifying all the TU & Go sub–tasks with a single IMU automatically providing information about variables such as: duration of sub–tasks, standing and sitting accelerations, rotation velocity of turning, number of steps during walking and turns, and the inclination degrees of the trunk during standing and sitting.

Towards an Automated Unsupervised Mobility Assessment for Older People Based on Inertial TUG Measurements

One of the most common assessments for the mobility of older people is the Timed Up and Go test (TUG). Due to its sensitivity regarding the indication of Parkinson’s disease (PD) or increased fall risk in elderly people, this assessment test becomes increasingly relevant, should be automated and should become applicable for unsupervised self-assessments to enable regular examinations of the functional status. With Inertial Measurement Units (IMU) being well suited for automated analyses, we evaluate an IMU-based analysis-system, which automatically detects the TUG execution via machine learning and calculates the test duration. as well as the duration of its single components. The complete TUG was classified with an accuracy of 96% via a rule-based model in a study with 157 participants aged over 70 years. A comparison between the TUG durations determined by IMU and criterion standard measurements (stopwatch and automated/ambient TUG (aTUG) system) showed significant correlations of 0.97 and 0.99, respectively. The classification of the instrumented TUG (iTUG)-components achieved accuracies over 96%, as well. Additionally, the system’s suitability for self-assessments was investigated within a semi-unsupervised situation where a similar movement sequence to the TUG was executed. This preliminary analysis confirmed that the self-selected speed correlates moderately with the speed in the test situation, but differed significantly from each other.

A Review on Accelerometry-Based Gait Analysis and Emerging Clinical Applications

Gait analysis continues to be an important technique for many clinical applications to diagnose and monitor certain diseases. Many mental and physical abnormalities cause measurable differences in a person's gait. Gait analysis has applications in sport, computer games, physical rehabilitation, clinical assessment, surveillance, human recognition, modeling, and many other fields. There are established methods using various sensors for gait analysis, of which accelerometers are one of the most often employed. Accelerometer sensors are generally more user friendly and less invasive. In this paper, we review research regarding accelerometer sensors used for gait analysis with particular focus on clinical applications. We provide a brief introduction to accelerometer theory followed by other popular sensing technologies. Commonly used gait phases and parameters are enumerated. The details of selecting the papers for review are provided. We also review several gait analysis software. Then we provide an extensive report of accelerometry-based gait analysis systems and applications, with additional emphasis on trunk accelerometry. We conclude this review with future research directions.

How Wearable Sensors Can Support Parkinson's Disease Diagnosis and Treatment: A Systematic Review

Background: Parkinson's disease (PD) is a common and disabling pathology that is characterized by both motor and non-motor symptoms and affects millions of people worldwide. The disease significantly affects quality of life of those affected. Many works in literature discuss the effects of the disease. The most promising trends involve sensor devices, which are low cost, low power, unobtrusive, and accurate in the measurements, for monitoring and managing the pathology.

OBJECTIVES

This review focuses on wearable devices for PD applications and identifies five main fields: early diagnosis, tremor, body motion analysis, motor fluctuations (ON-OFF phases), and home and long-term monitoring. The concept is to obtain an overview of the pathology at each stage of development, from the beginning of the disease to consider early symptoms, during disease progression with analysis of the most common disorders, and including management of the most complicated situations (i.e., motor fluctuations and long-term remote monitoring).

DATA SOURCES

The research was conducted within three databases: IEEE Xplore®, Science Direct®, and PubMed Central®, between January 2006 and December 2016.

STUDY ELIGIBILITY CRITERIA

Since 1,429 articles were found, accurate definition of the exclusion criteria and selection strategy allowed identification of the most relevant papers.

RESULTS

Finally, 136 papers were fully evaluated and included in this review, allowing a wide overview of wearable devices for the management of Parkinson's disease.

Convergent Validity of a Wearable Sensor System for Measuring Sub-Task Performance during the Timed Up-and-Go Test

Background: The timed-up-and-go test (TUG) is one of the most commonly used tests of physical function in clinical practice and for research outcomes. Inertial sensors have been used to parse the TUG test into its composite phases (rising, walking, turning, etc.), but have not validated this approach against an optoelectronic gold-standard, and to our knowledge no studies have published the minimal detectable change of these measurements. Methods: Eleven adults performed the TUG three times each under normal and slow walking conditions, and 3 m and 5 m walking distances, in a 12-camera motion analysis laboratory. An inertial measurement unit (IMU) with tri-axial accelerometers and gyroscopes was worn on the upper-torso. Motion analysis marker data and IMU signals were analyzed separately to identify the six main TUG phases: sit-to-stand, 1st walk, 1st turn, 2nd walk, 2nd turn, and stand-to-sit, and the absolute agreement between two systems analyzed using intra-class correlation (ICC, model 2) analysis. The minimal detectable change (MDC) within subjects was also calculated for each TUG phase. Results: The overall difference between TUG sub-tasks determined using 3D motion capture data and the IMU sensor data was <0.5 s. For all TUG distances and speeds, the absolute agreement was high for total TUG time and walk times (ICC > 0.90), but less for chair activity (ICC range 0.5–0.9) and typically poor for the turn time (ICC < 0.4). MDC values for total TUG time ranged between 2–4 s or 12–22% of the TUG time measurement. MDC of the sub-task times were higher proportionally, being 20–60% of the sub-task duration. Conclusions: We conclude that a commercial IMU can be used for quantifying the TUG phases with accuracy sufficient for clinical applications; however, the MDC when using inertial sensors is not necessarily improved over less sophisticated measurement tools.