Identifying freezing of gait in Parkinson's disease during freezing provoking tasks using waist-mounted accelerometry

Zach, Heidemarie; Janssen, A. Miranda L.; Snijders, Anke H.; Delval, Arnaud; Ferraye, Murielle Ursulla; Auff, Eduard; Weerdesteyn, Vivian; Bloem, Bastiaan R.; Nonnekes, Jorik

doi:10.1016/j.parkreldis.2015.09.051

Cited by 73 publications

(82 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The first half of the pool are 122 candidates, extracted using seven previously published features (i.e., average, standard deviation, variance, median, entropy, energy, power and FI as found in [12], [13], [14], [16], [17], [18], [19], [20], [23]) and our four aforementioned new features. We apply these eleven extraction functions to single and multiple inputs.…”

Section: B Feature Extraction 1) New Featuresmentioning

confidence: 99%

“…8Hz) divided by the power in the locomotor band (0.5 ! 3Hz) [12], has been used to build FoG detectors [12], [13], [16], [17], [18], [19], [20].…”

mentioning

confidence: 99%

“…Another work using Pearson's correlation introduced a rulebased 5-class classifier for strides including two classes for FoG with tremor and complete motor block [22]. As these reports were based on separate specific channels and several contexts, we compare our work with following similar studies: simple thresholding techniques [12], [13], [16], [19], [20] and supervised/semi-supervised learning classifiers [17], [18].…”

mentioning

confidence: 99%

“…Apart from FI, several features from accelerometer data (e.g., average, standard deviation, variance, median, entropy, energy, and power) have been proposed for FoG detectors [12], [13], [14], [16], [17], [18], [19], [20], [23]. Advanced statistical techniques to assess gait of human in general (e.g., postural control) can be found in a comprehensive feature investigation [24], however the work was concerned with 3D motion analysis for trajectory data using a single accelerometer at the lumbar.…”

mentioning

confidence: 99%

“…Using the vertical axis of an ankle sensor, a global threshold FI of 2.3 with 6s windowing was suggested in [12], then another global FI of 3 with 7s windowing was reported in [19]. By examining the same three locations of sensors with [19], reference [20] selected a different choice for the global FI of 1.4 (2s windows and the dorsoventral direction of the lumbar sensor). Model learning based classifiers have worked well for subject-dependent or group-dependent settings [17], [18].…”

mentioning

confidence: 99%

See 4 more Smart Citations

Freezing of Gait Detection in Parkinson's Disease: A Subject-Independent Detector Using Anomaly Scores

2017

IEEE Trans. Biomed. Eng.

View full text Add to dashboard Cite

Abstract-Freezing of gait (FoG) is common in Parkinsonian gait and strongly relates to falls. Current clinical FoG assessments are patients' self-report diaries and experts' manual video analysis. Both are subjective and yield moderate reliability. Existing detection algorithms have been predominantly designed in subject-dependent settings. In this work, we aim to develop an automated FoG detector for subject-independent. After extracting highly relevant features, we apply anomaly detection techniques to detect FoG events. Specifically, feature selection is performed using correlation and clusterability metrics. From a list of 244 feature candidates, 36 candidates were selected using saliency and robustness criteria. We develop an anomaly score detector (ASD) with adaptive thresholding to identify FoG events. Then, using accuracy metrics, we reduce the feature list to seven candidates. Our novel multi-channel freezing index was the most selective across all window sizes, achieving sensitivity (specificity) of 96% (79%). On the other hand, freezing index from the vertical axis was the best choice for a single input, achieving sensitivity (specificity) of 94% (84%) for ankle and 89% (94%) for back sensors. Our subject-independent method is not only significantly more accurate than those previously reported, but also uses a much smaller window (e.g., 3s vs. 7.5s) and/or lower tolerance (e.g., 0.4s vs. 2s).

show abstract

Section: B Feature Extraction 1) New Featuresmentioning

confidence: 99%

“…8Hz) divided by the power in the locomotor band (0.5 ! 3Hz) [12], has been used to build FoG detectors [12], [13], [16], [17], [18], [19], [20].…”

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Freezing of Gait Detection in Parkinson's Disease: A Subject-Independent Detector Using Anomaly Scores

2017

IEEE Trans. Biomed. Eng.

View full text Add to dashboard Cite

show abstract

Clinical and methodological challenges for assessing freezing of gait: Future perspectives

Bloem²,

et al. 2019

Self Cite

View full text Add to dashboard Cite

A BS TRACT: Freezing of gait, defined as sudden and usually brief episodes of inability to produce effective stepping, often results in falls and is both disabling and common in parkinsonism. In this narrative review, sprung from the 2nd International Workshop on freezing of gait in Leuven, we summarize the latest insights into clinical and methodological challenges for assessing freezing of gait. We also highlight the role of emerging wearable technology to improve the management of this debilitating symptom.

show abstract

Dual Task Turning in Place: A Reliable, Valid, and Responsive Outcome Measure of Freezing of Gait

et al. 2021

View full text Add to dashboard Cite

Background Freezing of gait (FOG) is a complex symptom in Parkinson's disease (PD) that is both elusive to elicit and varied in its presentation. These complexities present a challenge to measuring FOG in a sensitive and reliable way, precluding therapeutic advancement. Objective We investigated the reliability, validity, and responsiveness of manual video annotations of the turning‐in‐place task and compared it to the sensor‐based FOG ratio. Methods Forty‐five optimally medicated people with PD and FOG performed rapid alternating 360° turns without and with an auditory stroop dual task, thrice over two consecutive days. The tasks were video recorded, and inertial sensors were placed on the lower back and shins. Interrater reliability between three raters, criterion validity with self‐reported FOG, and responsiveness to single‐session split‐belt treadmill (SBT) training were investigated and contrasted with the sensor‐based FOG ratio. Results Visual ratings showed excellent agreement between raters for the percentage time frozen (%TF) (ICC [intra‐class correlation coefficient] = 0.99), the median duration of a FOG episode (ICC = 0.90), and the number of FOG episodes (ICC = 0.86). Dual tasking improved the sensitivity and validity of visual FOG ratings resulting in increased FOG detection, criterion validity with self‐reported FOG ratings, and responsiveness to a short SBT intervention. The sensor‐based FOG ratio, on the contrary, showed complex FOG presentation‐contingent relationships with visual and self‐reported FOG ratings and limited responsiveness to SBT training. Conclusions Manual video annotations of FOG during dual task turning in place generate reliable, valid, and sensitive outcomes for investigating therapeutic effects on FOG. © 2021 International Parkinson and Movement Disorder Society

show abstract

Identifying freezing of gait in Parkinson's disease during freezing provoking tasks using waist-mounted accelerometry

Cited by 73 publications

References 18 publications

Freezing of Gait Detection in Parkinson's Disease: A Subject-Independent Detector Using Anomaly Scores

Freezing of Gait Detection in Parkinson's Disease: A Subject-Independent Detector Using Anomaly Scores

Clinical and methodological challenges for assessing freezing of gait: Future perspectives

Dual Task Turning in Place: A Reliable, Valid, and Responsive Outcome Measure of Freezing of Gait

Contact Info

Product

Resources

About