Article
Chemistry, Analytical
Sebastian Rutkowski, Joren Buekers, Anna Rutkowska, Blazej Cieslik, Jan Szczegielniak
Summary: A study of 13 patients eligible for pulmonary rehabilitation found no significant difference in physical activity levels between training days and days off, as measured by the SenseWear armband. Using physical activity sensors may help motivate patients with COPD to be more active and improve health outcomes.
Article
Health Care Sciences & Services
Yingying Hao, Xiao-Kai Ma, Zheng Zhu, Zhen-Bo Cao
Summary: The study aimed to determine the validity of 11 commercially available wrist-wearable activity devices for monitoring total steps and total 24-hour total energy expenditure in healthy adolescents under simulated free-living conditions, with the conclusion that the Bong 2s had the best accuracy for estimating TEE and total steps. Further research is needed to examine the validity of these devices in different types of physical activities under real-world conditions.
JMIR MHEALTH AND UHEALTH
(2021)
Article
Chemistry, Multidisciplinary
Nora Alhammad, Hmood Al-Dossari
Summary: This study introduces a dynamic segmentation method for physical activity recognition during rehabilitation, which outperforms the traditional sliding window approach in terms of accuracy and model robustness.
APPLIED SCIENCES-BASEL
(2021)
Article
Engineering, Electrical & Electronic
Mahdi Pedram, Ramesh Kumar Sah, Seyed Ali Rokni, Marjan Nourollahi, Hassan Ghasemzadeh
Summary: Advances in embedded systems have led to the integration of wearable sensors in health monitoring. However, due to the personalized nature of human movement and the limitations of embedded sensors, a resource-efficient framework is needed for real-time activity recognition.
IEEE SENSORS JOURNAL
(2022)
Article
Computer Science, Information Systems
Junjie Zhang, Yuanhao Liu, Hua Yuan
Summary: Human Activity Recognition (HAR) commonly utilizes wearable sensors to analyze time series data and recognize specific actions. Existing approaches face challenges in differentiating similar actions when employing a combination of convolutional and recurrent neural networks. To improve recognition accuracy, integrating a residual structure and layer normalization into a bidirectional long short-term memory network (BLSTM) is proposed here. Experimental results on three public datasets (UCI-HAR, WISDM, and KU-HAR) demonstrate significant overall recognition accuracies of 98.37%, 99.01%, and 97.89% respectively, validating the effectiveness of this method. The codebase implementing the described framework can be found at: https://github.com/lyh0625/1DCNN-ResBLSTM-Attention.
Review
Sport Sciences
Karl E. Friedl, David P. Looney
Summary: Measuring human motion is a valuable source of health and physiological information. This paper explores motion-based biomarkers, such as movement patterns and quantified physical activity, that can be assessed using practical measurement technologies and evolving models and algorithms. The quantification of physical activity has advanced, allowing for better estimates of energy expenditure and classifications of activity types and intensity durations. Specific gaits and movement patterns can predict injury risks and reflect mood status. Emerging wearable systems are improving the identification of movement disorders and the medical management of chronic diseases. These advancements contribute to the enhancement of quality of life, protection of health, and improvement of performance.
JOURNAL OF SCIENCE AND MEDICINE IN SPORT
(2023)
Review
Sport Sciences
Karl E. Friedl, David P. Looney
Summary: Measures of human motion provide valuable information about health and physiological status. Examples of motion-based biomarkers include patterns of movement, quantified physical activity, and characteristic gaits. Practical measurement technologies and evolving physiological models and algorithms have made it possible to assess these biomarkers, with access to motion data and contextual information. Quantifying physical activity now goes beyond step counts, with accurate estimates of energy expenditure useful for weight management and health outcomes. Specific gaits can predict injury risk, and mood status can be inferred from certain types of human movement. Wearable systems can aid in the early identification of movement disorders and improve the management of chronic diseases. Emerging technologies, such as GPS and accelerometers, enable associations between health outcomes and performance, leading to better predictive models and algorithms.
JOURNAL OF SCIENCE AND MEDICINE IN SPORT
(2023)
Article
Materials Science, Multidisciplinary
Yue Li, Zibin Zhao, Asmita Veronica, Sin Yu Yeung, I-Ming Hsing
Summary: Continuous electrocardiography (ECG) monitoring is necessary for preventing cardiac events and premature death. This study introduces a 12-lead ECG patch that overcomes issues with compliance, skin damage, and user discomfort in conventional ECG sensors. The patch is safe, stretchable, and skin-adherent, allowing for user-friendly ECG monitoring.
ADVANCED MATERIALS TECHNOLOGIES
(2023)
Article
Chemistry, Analytical
Mohamed Elshafei, Diego Elias Costa, Emad Shihab
Summary: This research investigates the impact of muscle fatigue on Human Activity Recognition (HAR) systems, using biceps concentration curls as an example. Findings show that fatigue prolongs completion time of later sets and decreases muscular endurance, leading to changes in data patterns and affecting the performance of subject-specific and cross-subject models. Feedforward Neural Network (FNN) exhibits the best performance in both types of models.
Article
Engineering, Electrical & Electronic
Sunder Ali Khowaja, Parus Khuwaja, Kapal Dev, Muhammad Aslam Jarwar
Summary: In this study, we propose a PROMPT-based physical health monitoring framework that tracks subjective human behavior and handles the intensity variations associated with inertial measurement units. Experimental analysis shows that the proposed method achieves 14.56% better accuracy compared to existing works. We also propose a generalized framework for healthcare applications using wearable sensors and the PROMPT method for its triage with physical health monitoring systems in the real world.
IEEE SENSORS JOURNAL
(2023)
Article
Health Care Sciences & Services
Anis Davoudi, Mamoun T. Mardini, David Nelson, Fahd Albinali, Sanjay Ranka, Parisa Rashidi, Todd M. Manini
Summary: This study compared the accuracy of categorizing physical activity types and estimating energy expenditure in older adults using different placements of accelerometer devices. The results showed that additional accelerometer devices slightly improved activity recognition accuracy and MET estimation, but for older adults, single accelerometers with appropriate placement were sufficient.
JMIR MHEALTH AND UHEALTH
(2021)
Article
Computer Science, Information Systems
Fei Luo, Salabat Khan, Yandao Huang, Kaishun Wu
Summary: A wide range of sensors is used in human activity recognition, generating large amounts of data during monitoring. Server-based and cloud-based computing require uploading all sensor data, leading to increased costs and latency. However, the development of edge computing addresses this problem by moving computation and data storage closer to the sensors instead of relying on central servers/clouds.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Chemistry, Analytical
Alejandro Sanchez Guinea, Mehran Sarabchian, Max Muehlhaeuser
Summary: This paper proposes an approach to transform time-series data into images for activity recognition, aiming to address the reliance on complex deep learning models in current methods. Extensive evaluations show that the proposed approach outperforms existing techniques in all cases, while being easy to implement and extend.
Review
Endocrinology & Metabolism
Jose Manuel Jurado-Castro, Mercedes Gil-Campos, Francisco Jesus Llorente-Cantarero
Summary: This review summarizes recent evidence and advances in the implementation and use of new tools for assessing physical activity in children. Technological advances provide new objective methods, but there is still no consensus on the most appropriate approach for analyzing the duration and intensity of physical activity in children.
CURRENT OPINION IN CLINICAL NUTRITION AND METABOLIC CARE
(2022)
Article
Engineering, Biomedical
Jonathan Camargo, Will Flanagan, Noel Csomay-Shanklin, Bharat Kanwar, Aaron Young
Summary: This study develops and optimizes a combined locomotion mode classifier and environmental parameter estimator using machine learning and wearable sensors. The system accurately classifies ambulation modes, estimates ramp incline, stair height, and walking speed with high accuracy. Mechanical sensors play a significant role in classification, with goniometers dominating incline and height estimation, and speed estimation primarily relying on inertial measurement units.
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING
(2021)
Article
Chemistry, Analytical
Rui Zhang, Oliver Amft
Editorial Material
Computer Science, Information Systems
Oliver Amft, Jesus Favela, Stephen Intille, Micro Musolesi, Vassilis Kostakos
IEEE PERVASIVE COMPUTING
(2020)
Correction
Chemistry, Analytical
Giovanni Schiboni, Juan Carlos Suarez, Rui Zhang, Oliver Amft
Article
Computer Science, Artificial Intelligence
Andreas Rowald, Oliver Amft
Summary: Neurostimulation strategies offer a potential effective and safe alternative for treating various disorders. However, our understanding of the underlying mechanisms is limited, and thus personalized computational tools like Digital Neuro Twins (DNTs) can significantly contribute to identifying optimal stimulation parameters.
FRONTIERS IN NEUROROBOTICS
(2022)
Article
Computer Science, Information Systems
Giovanni Schiboni, Celia Martin Vicario, Juan Carlos Suarez, Federico Cruciani, Oliver Amft
Summary: This paper investigates a context-adaptive sample acquisition strategy at sub-Nyquist sampling rate for wearable embedded sensor devices. The approach can be applied to compressive sensing frameworks to minimize sampling and transmission costs. The results show that the approach saves between 13 to 22 percent of energy while achieving similar pattern recognition performance and reconstruction error.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2022)
Article
Computer Science, Information Systems
Annalisa Baronetto, Luisa S. Graf, Sarah Fischer, Markus F. Neurath, Oliver Amft
Summary: We used pretrained and non-pretrained deep neural models to detect 10-second Bowel Sounds audio segments in continuous audio data streams. The best model for segment-based Bowel Sounds spotting was EfficientNet-B2 with an attention module, and pretrained models significantly improved the F1 score. Our segment-based approach reduced the amount of audio data to be reviewed by experts from 84 hours to 11 hours, reducing workload by approximately 87%.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Maximilian Reiser, Andreas Breidenassel, Oliver Amft
Summary: This study analyzes the influence of reflective photoplethysmography (PPG) sensor positioning relative to blood vessels. The results show that a symmetrical arrangement of the PPG sensor around the blood vessel yields the maximum AC signal, while incorrect positioning leads to a deterioration in signal quality. Additionally, blood has the most profound effect on signal quality, and the position of the blood vessel also affects the mean penetration depth.
2022 IEEE-EMBS INTERNATIONAL CONFERENCE ON BIOMEDICAL AND HEALTH INFORMATICS (BHI'22) JOINTLY ORGANISED WITH THE IEEE-EMBS INTERNATIONAL CONFERENCE ON WEARABLE AND IMPLANTABLE BODY SENSOR NETWORKS (BSN'22)
(2022)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Addythia Saphala, Rui Zhang, Trinh Nam Thai, Oliver Amft
Summary: In this study, non-contact sensing of temporalis muscle contraction in smart eyeglasses frames was investigated for detecting eating activity. The results demonstrated that this approach has the potential to accurately detect chewing sequences and eating events, making it a highly promising tool for automated dietary monitoring.
2022 IEEE-EMBS INTERNATIONAL CONFERENCE ON BIOMEDICAL AND HEALTH INFORMATICS (BHI'22) JOINTLY ORGANISED WITH THE IEEE-EMBS INTERNATIONAL CONFERENCE ON WEARABLE AND IMPLANTABLE BODY SENSOR NETWORKS (BSN'22)
(2022)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Lena Uhlenberg, Oliver Amft
Summary: This study presents a framework for synthesising body-worn inertial sensor data based on personalised body surface and biomechanical models. By aligning mesh armatures and creating skeletal models, acceleration and angular velocity data were simulated for daily activities and compared against physical IMU measurements, showing lower errors in surface models.
2022 IEEE-EMBS INTERNATIONAL CONFERENCE ON BIOMEDICAL AND HEALTH INFORMATICS (BHI'22) JOINTLY ORGANISED WITH THE IEEE-EMBS INTERNATIONAL CONFERENCE ON WEARABLE AND IMPLANTABLE BODY SENSOR NETWORKS (BSN'22)
(2022)
Meeting Abstract
Sport Sciences
Erik A. Willis, Derek Hales, Falon Smith, Regan Burney, Michelle C. Rzepka, Oliver Amft, Rachel Barr, Kelly R. Evenson, Micheal R. Kosorok, Dianne S. Ward
MEDICINE & SCIENCE IN SPORTS & EXERCISE
(2022)
Article
Sport Sciences
Erik A. Willis, Derek Hales, Falon T. Smith, Regan Burney, Helal M. El-Zaatari, Michelle C. Rzepka, Oliver Amft, Rachel Barr, Kelly R. Evenson, Michael R. Kosorok, Dianne S. Ward
Summary: The study found that wearable sensors clipped to a child's shirt or embedded into eyeglasses are feasible and acceptable wear methods in free-living settings.
TRANSLATIONAL JOURNAL OF THE AMERICAN COLLEGE OF SPORTS MEDICINE
(2022)
Proceedings Paper
Computer Science, Cybernetics
David Kopyto, Rui Zhang, Oliver Amft
Summary: This paper compares three onset detection algorithms for acoustic chewing cycle detection, finding that the beat tracking algorithm performs the best in dietary monitoring. After leave-one-participant-out cross validation, the algorithm achieved 83% F-measure performance.
IWSC'21: PROCEEDINGS OF THE 2021 ACM INTERNATIONAL SYMPOSIUM ON WEARABLE COMPUTERS
(2021)
Proceedings Paper
Computer Science, Cybernetics
Annalisa Baronetto, Lena Uhlenberg, Dominik Wassermann, Oliver Amft
Summary: The study proposes a simulation method for evaluating the performance of garment-embedded contact sensors through dynamic 3D human body models and smart garment design. Analysis of sensor displacement and performance during common Activities of Daily Living showed that males had higher sensor displacement and displacement variance compared to females, while females had higher distance variance compared to males.
IWSC'21: PROCEEDINGS OF THE 2021 ACM INTERNATIONAL SYMPOSIUM ON WEARABLE COMPUTERS
(2021)
Proceedings Paper
Materials Science, Biomaterials
Annalisa Baronetto, Dominik Wassermann, Oliver Amft
Summary: The study introduces a framework for automatically extracting body landmarks and measurements from 3D body scans, which is then used in designing smart garments. Through a series of steps, the algorithm can accurately extract key body landmarks for fitting smart garments.
2021 IEEE 17TH INTERNATIONAL CONFERENCE ON WEARABLE AND IMPLANTABLE BODY SENSOR NETWORKS (BSN)
(2021)
Article
Chemistry, Analytical
Giovanni Schiboni, Juan Carlos Suarez, Rui Zhang, Oliver Amft