A Systematic Literature Review of the Factors that Influence the Accuracy of Consumer Wearable Health Device Data

Mahloko, Lerato; Adebesin, Funmi

doi:10.1007/978-3-030-45002-1_9

Cited by 17 publications

(9 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous works searching for the cause(s) of variability in pedometer accuracy have tested many variables [ 21 ], including the location on the body the device is worn, carried, or used [ 22 , 23 , 24 , 25 , 26 , 27 ], the device model or step counting algorithm [ 23 , 24 , 28 , 29 , 30 , 31 ], walking speed [ 23 , 24 , 25 , 26 , 28 , 30 , 32 , 33 , 34 ], age [ 32 ], presence of gait aids [ 35 ], weight [ 32 ], type of surface [ 33 , 34 , 36 ], and distance [ 31 ]. These variables have all been found to influence pedometer accuracy, as seen in Table 1 .…”

Section: Related Workmentioning

confidence: 99%

Evaluating Pedometer Algorithms on Semi-Regular and Unstructured Gaits

Mattfeld

Jesch

Hoover

2021

Sensors

View full text Add to dashboard Cite

Pedometers are popular for counting steps as a daily measure of physical activity, however, errors as high as 96% have been reported in previous work. Many reasons for pedometer error have been studied, including walking speed, sensor position on the body and pedometer algorithm, demonstrating some differences in error. However, we hypothesize that the largest source of error may be due to differences in the regularity of gait during different activities. During some activities, gait tends to be regular and the repetitiveness of individual steps makes them easy to identify in an accelerometer signal. During other activities of everyday life, gait is frequently semi-regular or unstructured, which we hypothesize makes it difficult to identify and count individual steps. In this work, we test this hypothesis by evaluating the three most common types of pedometer algorithm on a new data set that varies the regularity of gait. A total of 30 participants were video recorded performing three different activities: walking a path (regular gait), conducting a within-building activity (semi-regular gait), and conducting a within-room activity (unstructured gait). Participants were instrumented with accelerometers on the wrist, hip and ankle. Collectively, 60,805 steps were manually annotated for ground truth using synchronized video. The main contribution of this paper is to evaluate pedometer algorithms when the consistency of gait changes to simulate everyday life activities other than exercise. In our study, we found that semi-regular and unstructured gaits resulted in 5–466% error. This demonstrates the need to evaluate pedometer algorithms on activities that vary the regularity of gait. Our dataset is publicly available with links provided in the introduction and Data Availability Statement.

show abstract

Section: Related Workmentioning

confidence: 99%

Evaluating Pedometer Algorithms on Semi-Regular and Unstructured Gaits

Mattfeld

Jesch

Hoover

2021

Sensors

View full text Add to dashboard Cite

show abstract

“…The use of electrodes as a method of capturing data is thought to be the most reliable way for sensors to capture data from the body. In comparison, other methods, including the use of sensor belts, wrist watches, or PPG (photoplethysmography), seem to provide less robust readings [ 74 , 75 ].…”

Section: State-of-the-art Literature Reviewmentioning

confidence: 99%

Relevance of Objective Measures in Psychiatric Disorders—Rest-Activity Rhythm and Psychophysiological Measures

Moon

Yang

Seon

et al. 2021

Curr Psychiatry Rep

View full text Add to dashboard Cite

Purpose of Review We present a review of recent methods of objective measurement in psychiatry and psychology with a focus on home monitoring and its utility in guiding treatment. Recent Findings For individualized diagnostics and treatment of insomnia, actigraphy can generate clinically useful graphical presentations of sleep timing and patterns. Psychophysiological measures may complement psychometrics by tracking parallel changes in physiological responses and emotional functioning, especially during therapy for trauma symptoms and emotion regulation. It seems that rather than defining universal cut-offs, an individualised range of variability could characterize treatment response. Summary Wearable actigraphy and psychophysiological sensors are promising devices to provide biofeedback and guide treatment. Use of feasible and reliable technology during experimental and clinical procedures may necessitate defining healthy and abnormal responses in different populations and pathological states. We present a “call for action” towards further collaborative work to enable large scale use of objective measures.

show abstract

“…These raw sensor data are recorded from 51 subjects with 18 and 25 years who performed 18 pre-defined physical activities. All subjects wear the smartwatch on their dominant hand while they are performing the activities [28,29]. These physical activities can be categorized into three main categories, i.e., non-hand-oriented activities, hand-oriented eating activities, and hand-oriented general activities.…”

Section: Data Collectionmentioning

confidence: 99%

Enhanced Hand-Oriented Activity Recognition Based on Smartwatch Sensor Data Using LSTMs

et al. 2020

View full text Add to dashboard Cite

The creation of the Internet of Things (IoT), along with the latest developments in wearable technology, has provided new opportunities in human activity recognition (HAR). The modern smartwatch offers the potential for data from sensors to be relayed to novel IoT platforms, which allow the constant tracking and monitoring of human movement and behavior. Recently, traditional activity recognition techniques have done research in advance by choosing machine learning methods such as artificial neural network, decision tree, support vector machine, and naive Bayes. Nonetheless, these conventional machine learning techniques depend inevitably on heuristically handcrafted feature extraction, in which human domain knowledge is normally limited. This work proposes a hybrid deep learning model called CNN-LSTM that employed Long Short-Term Memory (LSTM) networks for activity recognition with the Convolution Neural Network (CNN). The study makes use of HAR involving smartwatches to categorize hand movements. Using the study based on the Wireless Sensor Data Mining (WISDM) public benchmark dataset, the recognition abilities of the deep learning model can be accessed. The accuracy, precision, recall, and F-measure statistics are employed using the evaluation metrics to assess the recognition abilities of LSTM models proposed. The findings indicate that this hybrid deep learning model offers better performance than its rivals, where the achievement of 96.2% accuracy, while the f-measure is 96.3%, is obtained. The results show that the proposed CNN-LSTM can support an improvement of the performance of activity recognition.

show abstract

A Systematic Literature Review of the Factors that Influence the Accuracy of Consumer Wearable Health Device Data

Cited by 17 publications

References 44 publications

Evaluating Pedometer Algorithms on Semi-Regular and Unstructured Gaits

Evaluating Pedometer Algorithms on Semi-Regular and Unstructured Gaits

Relevance of Objective Measures in Psychiatric Disorders—Rest-Activity Rhythm and Psychophysiological Measures

Enhanced Hand-Oriented Activity Recognition Based on Smartwatch Sensor Data Using LSTMs

Contact Info

Product

Resources

About