Dataset Information

Wrist accelerometer shape feature derivation methods for assessing activities of daily living.

ABSTRACT: BACKGROUND:There has been an increasing interest in understanding the usefulness of wrist-based accelerometer data for physical activity (PA) assessment due to the ease of use and higher user compliance than other body placements. PA assessment studies have relied on machine learning methods which take accelerometer data in forms of variables, or feature vectors. METHODS:In this work, we introduce automated shape feature derivation methods to transform epochs of accelerometer data into feature vectors. As the first step, recurring patterns in the collected data are identified and placed in a codebook. Similarities between epochs of accelerometer data and codebook's patterns are the basis of feature calculations. In this paper, we demonstrate supervised and unsupervised approaches to learn codebooks. We evaluated these methods and compared them with the standard statistical measures for PA assessment. The experiments were performed on 146 participants who wore an ActiGraph GT3X+ accelerometer on the right wrist and performed 33 activities of daily living. RESULTS:Our evaluations show that the shape feature derivation methods were able to perform comparably with the standard wrist model (F1-score: 0.89) for identifying sedentary PAs (F1-scores of 0.86 and 0.85 for supervised and unsupervised methods, respectively). This was also observed for identifying locomotion activities (F1-scores: 0.87, 0.83, and 0.81 for the standard wrist, supervised, unsupervised models, respectively). All the wrist models were able to estimate energy expenditure required for PAs with low error (rMSE: 0.90, 0.93, and 0.90 for the standard wrist, supervised, and unsupervised models, respectively). CONCLUSION:The automated shape feature derivation methods offer insights into the performed activities by providing a summary of repeating patterns in the accelerometer data. Furthermore, they could be used as efficient alternatives (or additions) for manually engineered features, especially important for cases where the latter fail to provide sufficient information to machine learning methods for PA assessment.

SUBMITTER: Kheirkhahan M

PROVIDER: S-EPMC6290590 | biostudies-other | 2018 Dec

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Wrist accelerometer shape feature derivation methods for assessing activities of daily living.

Kheirkhahan Matin M Chakraborty Avirup A Wanigatunga Amal A AA Corbett Duane B DB Manini Todd M TM Ranka Sanjay S

BMC medical informatics and decision making 20181212 Suppl 4

<h4>Background</h4>There has been an increasing interest in understanding the usefulness of wrist-based accelerometer data for physical activity (PA) assessment due to the ease of use and higher user compliance than other body placements. PA assessment studies have relied on machine learning methods which take accelerometer data in forms of variables, or feature vectors.<h4>Methods</h4>In this work, we introduce automated shape feature derivation methods to transform epochs of accelerometer data ...[more]

PMID: 30537957

Similar Datasets

Project description:Background: Huntington's disease (HD) leads to altered gait patterns and reduced daily-living physical activity. Accurate measurement of daily-living walking that takes into account involuntary movements (e.g. chorea) is needed. Objective: To evaluate daily-living gait quantity and quality in HD, taking into account irregular movements. Methods: Forty-two individuals with HD and fourteen age-matched non-HD peers completed clinic-based assessments and a standardized laboratory-based circuit of functional activities, wearing inertial measurement units on the wrists, legs, and trunk. These activities were used to train and test an algorithm for the automated detection of walking. Subsequently, 29 HD participants and 22 age-matched non-HD peers wore a tri-axial accelerometer on their non-dominant wrist for 7 days. Measures included gait quantity (e.g., steps per day), gait quality (e.g., regularity) metrics, and percentage of walking bouts with irregular movements. Results: Measures of daily-living gait quantity including step counts, walking time and bouts per day were similar in HD participants and non-HD peers (p > 0.05). HD participants with higher clinician-rated upper body chorea had a greater percentage of walking bouts with irregular movements compared to those with lower chorea (p = 0.060) and non-HD peers (p < 0.001). Even after accounting for irregular movements, within-bout walking consistency was lower in HD participants compared to non-HD peers (p < 0.001), while across-bout variability of these measures was higher (p < 0.001). Many of the daily-living measures were associated with disease-specific measures of motor function. Conclusions: Results suggest that a wrist-worn accelerometer can be used to evaluate the quantity and quality of daily-living gait in people with HD, while accounting for the influence of irregular (choreic-like) movements, and that gait features related to within- and across-bout consistency markedly differ in individuals with HD and non-HD peers.

Project description:Although accelerometry data are widely utilized to estimate physical activity and sedentary behavior among children age 3 years or older, for toddlers age 1 and 2 year(s), accelerometry data recorded during such behaviors have been far less examined. In particular, toddler's unique behaviors, such as riding in a stroller or being carried by an adult, have not yet been examined. The objective of this study was to describe accelerometry signal outputs recorded during participation in nine types of behaviors (i.e., running, walking, climbing up/down, crawling, riding a ride-on toy, standing, sitting, riding in a stroller/wagon, and being carried by an adult) among toddlers. Twenty-four toddlers aged 13 to 35 months (50% girls) performed various prescribed behaviors during free play in a commercial indoor playroom while wearing ActiGraph wGT3X-BT accelerometers on a hip and a wrist. Participants' performances were video-recorded. Based on the video data, accelerometer data were annotated with behavior labels to examine accelerometry signal outputs while performing the nine types of behaviors. Accelerometer data collected during 664 behavior assessments from the 21 participants were used for analysis. Hip vertical axis counts for walking were low (median = 49 counts/5 s). They were significantly lower than those recorded while a toddler was "carried" by an adult (median = 144 counts/5 s; p < 0.01). While standing, sitting, and riding in a stroller, very low hip vertical axis counts were registered (median ≤ 5 counts/5 s). Although wrist vertical axis and vector magnitude counts for "carried" were not higher than those for walking, they were higher than the cut-points for sedentary behaviors. Using various accelerometry signal features, machine learning techniques showed 89% accuracy to differentiate the "carried" behavior from ambulatory movements such as running, walking, crawling, and climbing. In conclusion, hip vertical axis counts alone may be unable to capture walking as physical activity and "carried" as sedentary behavior among toddlers. Machine learning techniques that utilize additional accelerometry signal features could help to recognize behavior types, especially to differentiate being "carried" from ambulatory movements.

Project description:Objectives:To evaluate application of the Barthel Index (BI) in assessing basic activities of daily living (ADL) of patients with dementia using Rasch analysis. Design:A multi-country cross-sectional study. Setting and Participants:Nineteen long-term care facilities located in China, Japan, South Korea, and Thailand. A total of 644 patients with dementia were included. Methods:Unidimensionality, global and item fit, local dependence, person-item targeting, threshold disordering, and differential item functioning (DIF) were examined. Negative correlations between scores for DIF items and Neuropsychiatric Inventory Nursing Home version (NPI-NH) were evaluated. Results:Item reliability (1.0) and person reliability (.88) were acceptable. The Rasch dimension explained 72.9% of the variance (Eigenvalue = 27), while the first contrast explained 6.6% (Eigenvalue = 2.4). The "mobility" was misfitting to the Rasch model (infit mean square = 1.86). The overall difficulty of the BI exceeded patients' ability (person location = -2.27 logits). The "stairs climbing" and "mobility" showed narrow category thresholds (< 1.4 logits). The location of "controlling bladder" and "toilet use" overlapped. Removing "stairs climbing", collapsing categories with narrow threshold widths in "mobility", and combining "controlling bowel" and "controlling bladder" into one item, improved unidimensionality, and item fit of the scale. Only three items ("grooming", "dressing", and "toilet use") were free from DIF across countries. The scores for "feeding" were negatively related to scores for "disinhibition" (r = -0.46, P < 0.01), and scores for "controlling bowel" were negatively related to scores for "disinhibition" (r = -0.44, P < 0.01), "agitation" (r = -0.32, P < 0.05), and "aggression" (r = -0.27, P < 0.01) in Japanese samples. Conclusions and Implications:The performance of the BI for assessing patients with dementia might be compromised by misfit items, person-item mistargeting, measurement gaps, redundant items, narrow threshold width, and item bias. Mobility ability might not be helpful for determining capability of basic ADL in the patients. Comparisons of BI scores between countries should be undertaken with caution due to item bias. Neuropsychiatric symptoms might interact with basic ADL abilities of the patients. We will not suggest using the instrument in patients with dementia, without future refining to improve its performance.

Dataset Information

Wrist accelerometer shape feature derivation methods for assessing activities of daily living.

Publications

Wrist accelerometer shape feature derivation methods for assessing activities of daily living.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets