A Comprehensive Analysis on Wearable Acceleration Sensors in Human Activity Recognition

SemiHAR: Improving Semisupervised Human Activity Recognition via Multitask Learning

Semisupervised human activity recognition (SemiHAR) has attracted attention in recent years from various domains, such as digital health and ambient intelligence. Currently, it still faces two challenges. For one thing, discriminative features may exist among multiple sequences rather than a single sequence since activities are combinations of motions involving several body parts. For another thing, labeled data and unlabeled data suffer from distribution discrepancies due to the different behavior patterns or biological conditions of users. For that, we propose a novel SemiHAR method based on multitask learning. First, a dimension-based Markov transition field (DMTF) technique is designed to generate 2-D activity data for capturing the interactions among different dimensions. Second, we jointly consider the user recognition (UR) task and the activity recognition (AR) task to reduce the underlying discrepancy. In addition, a task relation learner (TRL) is introduced to dynamically learn task relations, which enables the primary AR task to exploit preferred knowledge from other secondary tasks. We theoretically analyze the proposed SemiHAR and provide a novel generalization result. Extensive experiments conducted on four real-world datasets demonstrate that SemiHAR outperforms other state-of-the-art methods.

Impact of SARS-CoV-2 Infection on Essential Tremor: A Retrospective Clinical and Kinematic Analysis

In the past few years, SARS-CoV-2 infection has substantially impacted public health. Alongside respiratory symptoms, some individuals have reported new neurological manifestations or a worsening of pre-existing neurological conditions. We previously documented two cases of essential tremor (ET) who experienced a deterioration in tremor following SARS-CoV-2 infection. However, the effects of SARS-CoV-2 on ET remain largely unexplored. This study aims to evaluate the impact of SARS-CoV-2 infection on a relatively broad sample of ET patients by retrospectively comparing their clinical and kinematic data collected before and after the exposure to SARS-CoV-2. We surveyed to evaluate the impact of SARS-CoV-2 infection on tremor features in ET. Subsequently, we retrospectively analysed clinical and kinematic data, including accelerometric recordings of postural and kinetic tremor. We included 36 ET patients (14 females with a mean age of 71.1 ± 10.6 years). Among the 25 patients who reported SARS-CoV-2 infection, 11 (44%) noted a subjective worsening of tremor. All patients reporting subjective tremor worsening also exhibited symptoms of long COVID, whereas the prevalence of these symptoms was lower (50%) in those without subjective exacerbation. The retrospective analysis of clinical data revealed a tremor deterioration in infected patients, which was not observed in non-infected patients. Finally, kinematic analysis revealed substantial stability of tremor features in both groups. The study highlighted a potential correlation between the SARS-CoV-2 infection and clinical worsening of ET. Long COVID contributes to a greater impact of tremor on the daily life of ET patients.

Automatic detection of passing and shooting in water polo using machine learning: a feasibility study

There is currently no efficient way to quantify overhead throwing volume in water polo. Therefore, this study aimed to test the feasibility of a method to detect passes and shots in water polo automatically using inertial measurement units (IMU) and machine-learning algorithms. Eight water polo players wore one IMU sensor on the wrist (dominant hand) and one on the sacrum during six practices each. Sessions were filmed with a video camera and manually tagged for individual shots or passes. Data were synchronised between video tagging and IMU sensors using a cross-correlation approach. Support vector machine (SVM) and artificial neural networks (ANN) were compared based on sensitivity and specificity for identifying shots and passes. A total of 7294 actions were identified during the training sessions, including 945 shots and 5361 passes. Using SVM, passes and shots together were identified with 94.4% (95%CI = 91.8-96.4) sensitivity and 93.6% (95%CI = 91.4-95.4) specificity. Using ANN yielded similar sensitivity (93.0% [95%CI = 90.1-95.1]) and specificity (93.4% [95%CI = 91.1 = 95.2]). The results suggest that this method of identifying overhead throwing motions with IMU has potential for future field applications. A set-up with one single sensor at the wrist can suffice to measure these activities in water polo.

DiamondNet: A Neural-Network-Based Heterogeneous Sensor Attentive Fusion for Human Activity Recognition

With the proliferation of intelligent sensors integrated into mobile devices, fine-grained human activity recognition (HAR) based on lightweight sensors has emerged as a useful tool for personalized applications. Although shallow and deep learning algorithms have been proposed for HAR problems in the past decades, these methods have limited capability to exploit semantic features from multiple sensor types. To address this limitation, we propose a novel HAR framework, DiamondNet, which can create heterogeneous multisensor modalities, denoise, extract, and fuse features from a fresh perspective. In DiamondNet, we leverage multiple 1-D convolutional denoising autoencoders (1-D-CDAEs) to extract robust encoder features. We further introduce an attention-based graph convolutional network to construct new heterogeneous multisensor modalities, which adaptively exploit the potential relationship between different sensors. Moreover, the proposed attentive fusion subnet, which jointly employs a global-attention mechanism and shallow features, effectively calibrates different-level features of multiple sensor modalities. This approach amplifies informative features and provides a comprehensive and robust perception for HAR. The efficacy of the DiamondNet framework is validated on three public datasets. The experimental results demonstrate that our proposed DiamondNet outperforms other state-of-the-art baselines, achieving remarkable and consistent accuracy improvements. Overall, our work introduces a new perspective on HAR, leveraging the power of multiple sensor modalities and attention mechanisms to significantly improve the performance.

Temporal localization of upper extremity bilateral synergistic coordination using wearable accelerometers

Background

The human upper extremity is characterized by inherent motor abundance, allowing a diverse array of tasks with agility and adaptability. Upper extremity functional limitations are a common sequela to Stroke, resulting in pronounced motor and sensory impairments in the contralesional arm. While many therapeutic interventions focus on rehabilitating the weaker arm, it is increasingly evident that it is necessary to consider bimanual coordination and motor control.

Methods

Participants were recruited to two groups differing in age (Group 1 (n = 10): 23.4 ± 2.9 years, Group 2 (n = 10): 55.9 ± 10.6 years) for an exploratory study on the use of accelerometry to quantify bilateral coordination. Three tasks featuring coordinated reaching were selected to investigate the acceleration of the upper arm, forearm, and hand during activities of daily living (ADLs). Subjects were equipped with acceleration and inclination sensors on each upper arm, each forearm, and each hand. Data was segmented in MATLAB to assess inter-limb and intra-limb coordination. Inter-limb coordination was indicated through dissimilarity indices and temporal locations of congruous movement between upper arm, forearm, or hand segments of the right and left limbs. Intra-limb coordination was likewise assessed between upper arm-forearm, upper arm-hand, and forearm-hand segment pairs of the dominant limb.

Findings

Acceleration data revealed task-specific movement features during the three distinct tasks. Groups demonstrated diminished similarity as task complexity increased. Groups differed significantly in the hand segments during the buttoning task, with Group 1 showing no coordination in the hand segments during buttoning, and strong coordination in reaching each button with the upper arm and forearm guiding extension. Group 2's dissimilarity scores and percentages of similarity indicated longer periods of inter-limb coordination, particularly towards movement completion. Group 1's dissimilarity scores and percentages of similarity indicated longer periods of intra-limb coordination, particularly in the coordination of the upper arm and forearm segments.

Interpretation

The Expanding Procrustes methodology can be applied to compute objective coordination scores using accessible and highly accurate wearable acceleration sensors. The findings of task duration, angular velocity, and peak roll angle are supported by previous studies finding older individuals to present with slower movements, reduced movement stability, and a reduction of laterality between the limbs. The theory of a shift towards ambidexterity with age is supported by the finding of greater inter-limb coordination in the group of subjects above the age of thirty-five. The group below the age of thirty was found to demonstrate longer periods of intra-limb coordination, with upper arm and forearm coordination emerging as a possible explanation for the demonstrated greater stability.

Clinical and kinematic characterization of parkinsonian soft signs in essential tremor

BACKGROUND

Subtle parkinsonian signs, i.e., rest tremor and bradykinesia, are considered soft signs for defining essential tremor (ET) plus.

OBJECTIVES

Our study aimed to further characterize subtle parkinsonian signs in a relatively large sample of ET patients from a clinical and neurophysiological perspective.

METHODS

We employed clinical scales and kinematic techniques to assess a sample of 82 ET patients. Eighty healthy controls matched for gender and age were also included. The primary focus of our study was to conduct a comparative analysis of ET patients (without any soft signs) and ET-plus patients with rest tremor and/or bradykinesia. Additionally, we investigated the asymmetry and side concordance of these soft signs.

RESULTS

In ET-plus patients with parkinsonian soft signs (56.10% of the sample), rest tremor was clinically observed in 41.30% of cases, bradykinesia in 30.43%, and rest tremor plus bradykinesia in 28.26%. Patients with rest tremor had more severe and widespread action tremor than other patients. Furthermore, we observed a positive correlation between the amplitude of action and rest tremor. Most ET-plus patients had an asymmetry of rest tremor and bradykinesia. There was no side concordance between these soft signs, as confirmed through both clinical examination and kinematic evaluation.

CONCLUSIONS

Rest tremor and bradykinesia are frequently observed in ET and are often asymmetric but not concordant. Our findings provide a better insight into the phenomenology of ET and suggest that the parkinsonian soft signs (rest tremor and bradykinesia) in ET-plus may originate from distinct pathophysiological mechanisms.

TCN-attention-HAR: human activity recognition based on attention mechanism time convolutional network

Wearable sensors are widely used in medical applications and human-computer interaction because of their portability and powerful privacy. Human activity identification based on sensor data plays a vital role in these fields. Therefore, it is important to improve the recognition performance of different types of actions. Aiming at the problems of insufficient time-varying feature extraction and gradient explosion caused by too many network layers, a time convolution network recognition model with attention mechanism (TCN-Attention-HAR) was proposed. The model effectively recognizes and emphasizes the key feature information. The ability of extracting temporal features from TCN (temporal convolution network) is improved by using the appropriate size of the receiver domain. In addition, attention mechanisms are used to assign higher weights to important information, enabling models to learn and identify human activities more effectively. The performance of the Open Data Set (WISDM, PAMAP2 and USC-HAD) is improved by 1.13%, 1.83% and 0.51%, respectively, compared with other advanced models, these results clearly show that the network model presented in this paper has excellent recognition performance. In the knowledge distillation experiment, the parameters of student model are only about 0.1% of those of teacher model, and the accuracy of the model has been greatly improved, and in the WISDM data set, compared with the teacher's model, the accuracy is 0.14% higher.

3D Knee and Hip Angle Estimation With Reduced Wearable IMUs via Transfer Learning During Yoga, Golf, Swimming, Badminton, and Dance

Wearable lower-limb joint angle estimation using a reduced inertial measurement unit (IMU) sensor set could enable quick, economical sports injury risk assessment and motion capture; however the vast majority of existing research requires a full IMU set attached to every related body segment and is implemented in only a single movement, typically walking. We thus implemented 3-dimensional knee and hip angle estimation with a reduced IMU sensor set during yoga, golf, swimming (simulated lower body swimming in a seated posture), badminton, and dance movements. Additionally, current deep-learning models undergo an accuracy drop when tested with new and unseen activities, which necessitates collecting large amounts of data for the new activity. However, collecting large datasets for every new activity is time-consuming and expensive. Thus, a transfer learning (TL) approach with long short-term memory neural networks was proposed to enhance the model's generalization ability towards new activities while minimizing the need for a large new-activity dataset. This approach could transfer the generic knowledge acquired from training the model in the source-activity domain to the target-activity domain. The maximum improvement in estimation accuracy (RMSE) achieved by TL is 23.6 degrees for knee flexion/extension and 22.2 degrees for hip flexion/extension compared to without TL. These results extend the application of motion capture with reduced sensor configurations to a broader range of activities relevant to injury prevention and sports training. Moreover, they enhance the capacity of data-driven models in scenarios where acquiring a substantial amount of training data is challenging.

Elderly and visually impaired indoor activity monitoring based on Wi-Fi and Deep Hybrid convolutional neural network

A drop in physical activity and a deterioration in the capacity to undertake daily life activities are both connected with ageing and have negative effects on physical and mental health. An Elderly and Visually Impaired Human Activity Monitoring (EV-HAM) system that keeps tabs on a person's routine and steps in if a change in behaviour or a crisis might greatly help an elderly person or a visually impaired. These individuals may find greater freedom with the help of an EVHAM system. As the backbone of human-centric applications like actively supported living and in-home monitoring for the elderly and visually impaired, an EVHAM system is essential. Big data-driven product design is flourishing in this age of 5G and the IoT. Recent advancements in processing power and software architectures have also contributed to the emergence and development of artificial intelligence (AI). In this context, the digital twin has emerged as a state-of-the-art technology that bridges the gap between the real and virtual worlds by evaluating data from several sensors using artificial intelligence algorithms. Although promising findings have been reported by Wi-Fi-based human activity identification techniques so far, their effectiveness is vulnerable to environmental variations. Using the environment-independent fingerprints generated from the Wi-Fi channel state information (CSI), we introduce Wi-Sense. This human activity identification system employs a Deep Hybrid convolutional neural network (DHCNN). The proposed system begins by collecting the CSI with a regular Wi-Fi Network Interface Controller. Wi-Sense uses the CSI ratio technique to lessen the effect of noise and the phase offset. The t- Distributed Stochastic Neighbor Embedding (t-SNE) is used to eliminate unnecessary data further. The data dimension is decreased, and the negative effects on the environment are eliminated in this process. The resulting spectrogram of the processed data exposes the activity's micro-Doppler fingerprints as a function of both time and location. These spectrograms are put to use in the training of a DHCNN. Based on our findings, EVHAM can accurately identify these actions 99% of the time.

Prediction of Physical Activity Patterns in Older Patients Rehabilitating After Hip Fracture Surgery: Exploratory Study

BACKGROUND

Building up physical activity is a highly important aspect in an older patient's rehabilitation process after hip fracture surgery. The patterns of physical activity during rehabilitation are associated with the duration of rehabilitation stay. Predicting physical activity patterns early in the rehabilitation phase can provide patients and health care professionals an early indication of the duration of rehabilitation stay as well as insight into the degree of patients' recovery for timely adaptive interventions.

OBJECTIVE

This study aims to explore the early prediction of physical activity patterns in older patients rehabilitating after hip fracture surgery at a skilled nursing home.

METHODS

The physical activity of patients aged ≥70 years with surgically treated hip fracture was continuously monitored using an accelerometer during rehabilitation at a skilled nursing home. Physical activity patterns were described in our previous study, and the 2 most common patterns were used in this study for pattern prediction: the upward linear pattern (n=15) and the S-shape pattern (n=23). Features from the intensity of physical activity were calculated for time windows with different window sizes of the first 5, 6, 7, and 8 days to assess the early rehabilitation moment in which the patterns could be predicted most accurately. Those features were statistical features, amplitude features, and morphological features. Furthermore, the Barthel Index, Fracture Mobility Score, Functional Ambulation Categories, and the Montreal Cognitive Assessment score were used as clinical features. With the correlation-based feature selection method, relevant features were selected that were highly correlated with the physical activity patterns and uncorrelated with other features. Multiple classifiers were used: decision trees, discriminant analysis, logistic regression, support vector machines, nearest neighbors, and ensemble classifiers. The performance of the prediction models was assessed by calculating precision, recall, and F1-score (accuracy measure) for each individual physical activity pattern. Furthermore, the overall performance of the prediction model was calculated by calculating the F1-score for all physical activity patterns together.

RESULTS

The amplitude feature describing the overall intensity of physical activity on the first day of rehabilitation and the morphological features describing the shape of the patterns were selected as relevant features for all time windows. Relevant features extracted from the first 7 days with a cosine k-nearest neighbor model reached the highest overall prediction performance (micro F1-score=1) and a 100% correct classification of the 2 most common physical activity patterns.

CONCLUSIONS

Continuous monitoring of the physical activity of older patients in the first week of hip fracture rehabilitation results in an early physical activity pattern prediction. In the future, continuous physical activity monitoring can offer the possibility to predict the duration of rehabilitation stay, assess the recovery progress during hip fracture rehabilitation, and benefit health care organizations, health care professionals, and patients themselves.

Development and Testing of a Data Capture Device for Use With Clinical Incentive Spirometers: Testing and Usability Study

BACKGROUND

The incentive spirometer is a basic and common medical device from which electronic health care data cannot be directly collected. As a result, despite numerous studies investigating clinical use, there remains little consensus on optimal device use and sparse evidence supporting its intended benefits such as prevention of postoperative respiratory complications.

OBJECTIVE

The aim of the study is to develop and test an add-on hardware device for data capture of the incentive spirometer.

METHODS

An add-on device was designed, built, and tested using reflective optical sensors to identify the real-time location of the volume piston and flow bobbin of a common incentive spirometer. Investigators manually tested sensor level accuracies and triggering range calibrations using a digital flowmeter. A valid breath classification algorithm was created and tested to determine valid from invalid breath attempts. To assess real-time use, a video game was developed using the incentive spirometer and add-on device as a controller using the Apple iPad.

RESULTS

In user testing, sensor locations were captured at an accuracy of 99% (SD 1.4%) for volume and 100% accuracy for flow. Median and average volumes were within 7.5% (SD 6%) of target volume sensor levels, and maximum sensor triggering values seldom exceeded intended sensor levels, showing a good correlation to placement on 2 similar but distinct incentive spirometer designs. The breath classification algorithm displayed a 100% sensitivity and a 99% specificity on user testing, and the device operated as a video game controller in real time without noticeable interference or delay.

CONCLUSIONS

An effective and reusable add-on device for the incentive spirometer was created to allow the collection of previously inaccessible incentive spirometer data and demonstrate Internet-of-Things use on a common hospital device. This design showed high sensor accuracies and the ability to use data in real-time applications, showing promise in the ability to capture currently inaccessible clinical data. Further use of this device could facilitate improved research into the incentive spirometer to improve adoption, incentivize adherence, and investigate the clinical effectiveness to help guide clinical care.

Reducing the Impact of Sensor Orientation Variability in Human Activity Recognition Using a Consistent Reference System

Sensor- orientation is a critical aspect in a Human Activity Recognition (HAR) system based on tri-axial signals (such as accelerations); different sensors orientations introduce important errors in the activity recognition process. This paper proposes a new preprocessing module to reduce the negative impact of sensor-orientation variability in HAR. Firstly, this module estimates a consistent reference system; then, the tri-axial signals recorded from sensors with different orientations are transformed into this consistent reference system. This new preprocessing has been evaluated to mitigate the effect of different sensor orientations on the classification accuracy in several state-of-the-art HAR systems. The experiments were carried out using a subject-wise cross-validation methodology over six different datasets, including movements and postures. This new preprocessing module provided robust HAR performance even when sudden sensor orientation changes were included during data collection in the six different datasets. As an example, for the WISDM dataset, sensors with different orientations provoked a significant reduction in the classification accuracy of the state-of-the-art system (from 91.57 ± 0.23% to 89.19 ± 0.26%). This important reduction was recovered with the proposed algorithm, increasing the accuracy to 91.46 ± 0.30%, i.e., the same result obtained when all sensors had the same orientation.

Longitudinal Studies of Wearables in Patients with Diabetes: Key Issues and Solutions

Glucose monitoring is key to the management of diabetes mellitus to maintain optimal glucose control whilst avoiding hypoglycemia. Non-invasive continuous glucose monitoring techniques have evolved considerably to replace finger prick testing, but still require sensor insertion. Physiological variables, such as heart rate and pulse pressure, change with blood glucose, especially during hypoglycemia, and could be used to predict hypoglycemia. To validate this approach, clinical studies that contemporaneously acquire physiological and continuous glucose variables are required. In this work, we provide insights from a clinical study undertaken to study the relationship between physiological variables obtained from a number of wearables and glucose levels. The clinical study included three screening tests to assess neuropathy and acquired data using wearable devices from 60 participants for four days. We highlight the challenges and provide recommendations to mitigate issues that may impact the validity of data capture to enable a valid interpretation of the outcomes.

In-Bed Posture Classification Using Deep Neural Network

In-bed posture monitoring has become a prevalent area of research to help minimize the risk of pressure sore development and to increase sleep quality. This paper proposed 2D and 3D Convolutional Neural Networks, which are trained on images and videos of an open-access dataset consisting of 13 subjects' body heat maps captured from a pressure mat in 17 positions, respectively. The main goal of this paper is to detect the three main body positions: supine, left, and right. We compare the use of image and video data through 2D and 3D models in our classification. Since the dataset was imbalanced, three strategies were evaluated, i.e., down sampling, over sampling, and class weights. The best 3D model achieved accuracies of 98.90 ± 1.05% and 97.80 ± 2.14% for 5-fold and leave-one-subject-out (LOSO) cross validations, respectively. To compare the 3D model with 2D, four pre-trained 2D models were evaluated, where the best-performing model was the ResNet-18 with accuracies of 99.97 ± 0.03% for 5-fold and 99.62 ± 0.37% for LOSO. The proposed 2D and 3D models provided promising results for in-bed posture recognition and can be used in the future to further distinguish postures into more detailed subclasses. The outcome of this study can be used to remind caregivers at hospitals and long-term care facilitiesto reposition their patients if they do not reposition themselves naturally to prevent pressure ulcers. In addition, the evaluation of body postures and movements during sleep can help caregivers understand sleep quality.

A Multimodal Data Processing System for LiDAR-Based Human Activity Recognition

Increasingly, the task of detecting and recognizing the actions of a human has been delegated to some form of neural network processing camera or wearable sensor data. Due to the degree to which the camera can be affected by lighting and wearable sensors scantiness, neither one modality can capture the required data to perform the task confidently. That being the case, range sensors, like light detection and ranging (LiDAR), can complement the process to perceive the environment more robustly. Most recently, researchers have been exploring ways to apply convolutional neural networks to 3-D data. These methods typically rely on a single modality and cannot draw on information from complementing sensor streams to improve accuracy. This article proposes a framework to tackle human activity recognition by leveraging the benefits of sensor fusion and multimodal machine learning. Given both RGB and point cloud data, our method describes the activities being performed by subjects using regions with a convolutional neural network (R-CNN) and a 3-D modified Fisher vector network. Evaluated on a custom captured multimodal dataset demonstrates that the model outputs remarkably accurate human activity classification (90%). Furthermore, this framework can be used for sports analytics, understanding social behavior, surveillance, and perhaps most notably by autonomous vehicles (AVs) to data-driven decision-making policies in urban areas and indoor environments.

Context-Aware Edge-Based AI Models for Wireless Sensor Networks-An Overview

Recent advances in sensor technology are expected to lead to a greater use of wireless sensor networks (WSNs) in industry, logistics, healthcare, etc. On the other hand, advances in artificial intelligence (AI), machine learning (ML), and deep learning (DL) are becoming dominant solutions for processing large amounts of data from edge-synthesized heterogeneous sensors and drawing accurate conclusions with better understanding of the situation. Integration of the two areas WSN and AI has resulted in more accurate measurements, context-aware analysis and prediction useful for smart sensing applications. In this paper, a comprehensive overview of the latest developments in context-aware intelligent systems using sensor technology is provided. In addition, it also discusses the areas in which they are used, related challenges, motivations for adopting AI solutions, focusing on edge computing, i.e., sensor and AI techniques, along with analysis of existing research gaps. Another contribution of this study is the use of a semantic-aware approach to extract survey-relevant subjects. The latter specifically identifies eleven main research topics supported by the articles included in the work. These are analyzed from various angles to answer five main research questions. Finally, potential future research directions are also discussed.

Sense and Learn: Recent Advances in Wearable Sensing and Machine Learning for Blood Glucose Monitoring and Trend-Detection

Diabetes mellitus is characterized by elevated blood glucose levels, however patients with diabetes may also develop hypoglycemia due to treatment. There is an increasing demand for non-invasive blood glucose monitoring and trends detection amongst people with diabetes and healthy individuals, especially athletes. Wearable devices and non-invasive sensors for blood glucose monitoring have witnessed considerable advances. This review is an update on recent contributions utilizing novel sensing technologies over the past five years which include electrocardiogram, electromagnetic, bioimpedance, photoplethysmography, and acceleration measures as well as bodily fluid glucose sensors to monitor glucose and trend detection. We also review methods that use machine learning algorithms to predict blood glucose trends, especially for high risk events such as hypoglycemia. Convolutional and recurrent neural networks, support vector machines, and decision trees are examples of such machine learning algorithms. Finally, we address the key limitations and challenges of these studies and provide recommendations for future work.

Wearable sensors during drawing tasks to measure the severity of essential tremor

Commonly used methods to assess the severity of essential tremor (ET) are based on clinical observation and lack objectivity. This study proposes the use of wearable accelerometer sensors for the quantitative assessment of ET. Acceleration data was recorded by inertial measurement unit (IMU) sensors during sketching of Archimedes spirals in 17 ET participants and 18 healthy controls. IMUs were placed at three points (dorsum of hand, posterior forearm, posterior upper arm) of each participant's dominant arm. Movement disorder neurologists who were blinded to clinical information scored ET patients on the Fahn-Tolosa-Marin rating scale (FTM) and conducted phenotyping according to the recent Consensus Statement on the Classification of Tremors. The ratio of power spectral density of acceleration data in 4-12 Hz to 0.5-4 Hz bands and the total duration of the action were inputs to a support vector machine that was trained to classify the ET subtype. Regression analysis was performed to determine the relationship of acceleration and temporal data with the FTM scores. The results show that the sensor located on the forearm had the best classification and regression results, with accuracy of 85.71% for binary classification of ET versus control. There was a moderate to good correlation (r2 = 0.561) between FTM and a combination of power spectral density ratio and task time. However, the system could not accurately differentiate ET phenotypes according to the Consensus classification scheme. Potential applications of machine-based assessment of ET using wearable sensors include clinical trials and remote monitoring of patients.

Toward the Personalization of Biceps Fatigue Detection Model for Gym Activity: An Approach to Utilize Wearables' Data from the Crowd

Nowadays, wearables-based Human Activity Recognition (HAR) systems represent a modern, robust, and lightweight solution to monitor athlete performance. However, user data variability is a problem that may hinder the performance of HAR systems, especially the cross-subject HAR models. Such a problem may have a lesser effect on the subject-specific model because it is a tailored model that serves a specific user; hence, data variability is usually low, and performance is often high. However, such a performance comes with a high cost in data collection and processing per user. Therefore, in this work, we present a personalized model that achieves higher performance than the cross-subject model while maintaining a lower data cost than the subject-specific model. Our personalization approach sources data from the crowd based on similarity scores computed between the test subject and the individuals in the crowd. Our dataset consists of 3750 concentration curl repetitions from 25 volunteers with ages and BMI ranging between 20-46 and 24-46, respectively. We compute 11 hand-crafted features and train 2 personalized AdaBoost models, Decision Tree (AdaBoost-DT) and Artificial Neural Networks (AdaBoost-ANN), using data from whom the test subject shares similar physical and single traits. Our findings show that the AdaBoost-DT model outperforms the cross-subject-DT model by 5.89%, while the AdaBoost-ANN model outperforms the cross-subject-ANN model by 3.38%. On the other hand, at 50.0% less of the test subject's data consumption, our AdaBoost-DT model outperforms the subject-specific-DT model by 16%, while the AdaBoost-ANN model outperforms the subject-specific-ANN model by 10.33%. Yet, the subject-specific models achieve the best performances at 100% of the test subjects' data consumption.

A CSI-Based Multi-Environment Human Activity Recognition Framework

Passive human activity recognition (HAR) systems, in which no sensors are attached to the subject, provide great potentials compared to conventional systems. One of the recently used techniques showing tremendous potential is channel state information (CSI)-based HAR systems. In this work, we present a multi-environment human activity recognition system based on observing the changes in the CSI values of the exchanged wireless packets carried by OFDM subcarriers. In essence, we introduce a five-stage CSI-based human activity recognition approach. First, the acquired CSI values associated with each recorded activity instance are processed to remove the existing noise from the recorded data. A novel segmentation algorithm is then presented to identify and extract the portion of the signal that contains the activity. Next, the extracted activity segment is processed using the procedure proposed in the first stage. After that, the relevant features are extracted, and the important features are selected. Finally, the selected features are used to train a support vector machine (SVM) classifier to identify the different performed activities. To validate the performance of the proposed approach, we collected data in two different environments. In each of the environments, several activities were performed by multiple subjects. The performed experiments showed that our proposed approach achieved an average activity recognition accuracy of 91.27%.

A Novel Elderly Tracking System Using Machine Learning to Classify Signals from Mobile and Wearable Sensors

A health or activity monitoring system is the most promising approach to assisting the elderly in their daily lives. The increase in the elderly population has increased the demand for health services so that the existing monitoring system is no longer able to meet the needs of sufficient care for the elderly. This paper proposes the development of an elderly tracking system using the integration of multiple technologies combined with machine learning to obtain a new elderly tracking system that covers aspects of activity tracking, geolocation, and personal information in an indoor and an outdoor environment. It also includes information and results from the collaboration of local agencies during the planning and development of the system. The results from testing devices and systems in a case study show that the k-nearest neighbor (k-NN) model with k = 5 was the most effective in classifying the nine activities of the elderly, with 96.40% accuracy. The developed system can monitor the elderly in real-time and can provide alerts. Furthermore, the system can display information of the elderly in a spatial format, and the elderly can use a messaging device to request help in an emergency. Our system supports elderly care with data collection, tracking and monitoring, and notification, as well as by providing supporting information to agencies relevant in elderly care.

Transfer Learning Improves Accelerometer-Based Child Activity Recognition via Subject-Independent Adult-Domain Adaption

Wearable activity recognition can collate the type, intensity, and duration of each childs physical activity profile, which is important for exploring underlying adolescent health mechanisms. Traditional machine-learning-based approaches require large labeled data sets; however, child activity data sets are typically small and insufficient. Thus, we proposed a transfer learning approach that adapts adult-domain data to train a high-fidelity, subject-independent model for child activity recognition. Twenty children and twenty adults wore an accelerometer wristband while performing walking, running, sitting, and rope skipping activities. Activity classification accuracy was determined via the traditional machine learning approach without transfer learning and with the proposed subject-independent transfer learning approach. Results showed that transfer learning increased classification accuracy to 91.4% as compared to 80.6% without transfer learning. These results suggest that subject-independent transfer learning can improve accuracy and potentially reduce the size of the required child data sets to enable physical activity monitoring systems to be adopted more widely, quickly, and economically for children and provide deeper insights into injury prevention and health promotion strategies.

Recognizing Physical Activities for Spinal Cord Injury Rehabilitation Using Wearable Sensors

The research area of activity recognition is fast growing with diverse applications. However, advances in this field have not yet been used to monitor the rehabilitation of individuals with spinal cord injury. Noteworthily, relying on patient surveys to assess adherence can undermine the outcomes of rehabilitation. Therefore, this paper presents and implements a systematic activity recognition method to recognize physical activities applied by subjects during rehabilitation for spinal cord injury. In the method, raw sensor data are divided into fragments using a dynamic segmentation technique, providing higher recognition performance compared to the sliding window, which is a commonly used approach. To develop the method and build a predictive model, a machine learning approach was adopted. The proposed method was evaluated on a dataset obtained from a single wrist-worn accelerometer. The results demonstrated the effectiveness of the proposed method in recognizing all of the activities that were examined, and it achieved an overall accuracy of 96.86%.

Human Activity Recognition for People with Knee Osteoarthritis-A Proof-of-Concept

Clinicians lack objective means for monitoring if their knee osteoarthritis patients are improving outside of the clinic (e.g., at home). Previous human activity recognition (HAR) models using wearable sensor data have only used data from healthy people and such models are typically imprecise for people who have medical conditions affecting movement. HAR models designed for people with knee osteoarthritis have classified rehabilitation exercises but not the clinically relevant activities of transitioning from a chair, negotiating stairs and walking, which are commonly monitored for improvement during therapy for this condition. Therefore, it is unknown if a HAR model trained on data from people who have knee osteoarthritis can be accurate in classifying these three clinically relevant activities. Therefore, we collected inertial measurement unit (IMU) data from 18 participants with knee osteoarthritis and trained convolutional neural network models to identify chair, stairs and walking activities, and phases. The model accuracy was 85% at the first level of classification (activity), 89-97% at the second (direction of movement) and 60-67% at the third level (phase). This study is the first proof-of-concept that an accurate HAR system can be developed using IMU data from people with knee osteoarthritis to classify activities and phases of activities.

Development and validation of smartwatch-based activity recognition models for rigging crew workers on cable logging operations

Analysis of high-resolution inertial sensor and global navigation satellite system (GNSS) data collected by mobile and wearable devices is a relatively new methodology in forestry and safety research that provides opportunities for modeling work activities in greater detail than traditional time study analysis. The objective of this study was to evaluate whether smartwatch-based activity recognition models could quantify the activities of rigging crew workers setting and disconnecting log chokers on cable logging operations. Four productive cycle elements (travel to log, set choker, travel away, clear) were timed for choker setters and four productive cycle elements (travel to log, unhook, travel away, clear) were timed for chasers working at five logging sites in North Idaho. Each worker wore a smartwatch that recorded accelerometer data at 25 Hz. Random forest machine learning was used to develop predictive models that classified the different cycle elements based on features extracted from the smartwatch acceleration data using 15 sliding window sizes (1 to 15 s) and five window overlap levels (0%, 25%, 50%, 75%, and 90%). Models were compared using multiclass area under the Receiver Operating Characteristic (ROC) curve, or AUC. The best choker setter model was created using a 3-s window with 90% overlap and had sensitivity values ranging from 76.95% to 83.59% and precision values ranging from 41.42% to 97.08%. The best chaser model was created using a 1-s window with 90% overlap and had sensitivity values ranging from 71.95% to 82.75% and precision values ranging from 14.74% to 99.16%. These results have demonstrated the feasibility of quantifying forestry work activities using smartwatch-based activity recognition models, a basic step needed to develop real-time safety notifications associated with high-risk job functions and to advance subsequent, comparative analysis of health and safety metrics across stand, site, and work conditions.

Human Activity Recognition Based on Dynamic Active Learning

Activity of daily living is an important indicator of the health status and functional capabilities of an individual. Activity recognition, which aims at understanding the behavioral patterns of people, has increasingly received attention in recent years. However, there are still a number of challenges confronting the task. First, labelling training data is expensive and time-consuming, leading to limited availability of annotations. Secondly, activities performed by individuals have considerable variability, which renders the generally used supervised learning with a fixed label set unsuitable. To address these issues, we propose a dynamic active learning-based activity recognition method in this work. Different from traditional active learning methods which select samples based on a fixed label set, the proposed method not only selects informative samples from known classes, but also dynamically identifies new activities which are not included in the predefined label set. Starting with a classifier that has access to a limited number of labelled samples, we iteratively extend the training set with informative labels by fully considering the uncertainty, diversity and representativeness of samples, based on which better-informed classifiers can be trained, further reducing the annotation cost. We evaluate the proposed method on two synthetic datasets and two existing benchmark datasets. Experimental results demonstrate that our method not only boosts the activity recognition performance with considerably reduced annotation cost, but also enables adaptive daily activity analysis allowing the presence and detection of novel activities and patterns.

Dynamic Segmentation for Physical Activity Recognition Using a Single Wearable Sensor

Data segmentation is an essential process in activity recognition when using machine learning techniques. Previous studies on physical activity recognition have mostly relied on the sliding window approach for segmentation. However, choosing a fixed window size for multiple activities with different durations may affect recognition accuracy, especially when the activities belong to the same category (i.e., dynamic or static). This paper presents and verifies a new method for dynamic segmentation of physical activities performed during the rehabilitation of individuals with spinal cord injuries. To adaptively segment the raw data, signal characteristics are analyzed to determine the suitable type of boundaries. Then, the algorithm identifies the time boundaries to represent the start- and endpoints of each activity. To verify the method and build a predictive model, an experiment was conducted in which data were collected using a single wrist-worn accelerometer sensor. The experimental results were compared with the sliding window approach, indicating that the proposed method outperformed the sliding window approach in terms of overall accuracy, which exceeded 5%, as well as model robustness. The results also demonstrated efficient physical activity segmentation using the proposed method, resulting in high classification performance for all activities considered.

Objective Quantification of In-Hospital Patient Mobilization after Cardiac Surgery Using Accelerometers: Selection, Use, and Analysis

Cardiac surgery patients infrequently mobilize during their hospital stay. It is unclear for patients why mobilization is important, and exact progress of mobilization activities is not available. The aim of this study was to select and evaluate accelerometers for objective qualification of in-hospital mobilization after cardiac surgery. Six static and dynamic patient activities were defined to measure patient mobilization during the postoperative hospital stay. Device requirements were formulated, and the available devices reviewed. A triaxial accelerometer (AX3, Axivity) was selected for a clinical pilot in a heart surgery ward and placed on both the upper arm and upper leg. An artificial neural network algorithm was applied to classify lying in bed, sitting in a chair, standing, walking, cycling on an exercise bike, and walking the stairs. The primary endpoint was the daily amount of each activity performed between 7 a.m. and 11 p.m. The secondary endpoints were length of intensive care unit stay and surgical ward stay. A subgroup analysis for male and female patients was planned. In total, 29 patients were classified after cardiac surgery with an intensive care unit stay of 1 (1 to 2) night and surgical ward stay of 5 (3 to 6) nights. Patients spent 41 (20 to 62) min less time in bed for each consecutive hospital day, as determined by a mixed-model analysis (p < 0.001). Standing, walking, and walking the stairs increased during the hospital stay. No differences between men (n = 22) and women (n = 7) were observed for all endpoints in this study. The approach presented in this study is applicable for measuring all six activities and for monitoring postoperative recovery of cardiac surgery patients. A next step is to provide feedback to patients and healthcare professionals, to speed up recovery.

Acceleration Feature Extraction of Human Body Based on Wearable Devices

Wearable devices used for human body monitoring has broad applications in smart home, sports, security and other fields. Wearable devices provide an extremely convenient way to collect a large amount of human motion data. In this paper, the human body acceleration feature extraction method based on wearable devices is studied. Firstly, Butterworth filter is used to filter the data. Then, in order to ensure the extracted feature value more accurately, it is necessary to remove the abnormal data in the source. This paper combines Kalman filter algorithm with a genetic algorithm and use the genetic algorithm to code the parameters of the Kalman filter algorithm. We use Standard Deviation (SD), Interval of Peaks (IoP) and Difference between Adjacent Peaks and Troughs (DAPT) to analyze seven kinds of acceleration. At last, SisFall data set, which is a globally available data set for study and experiments, is used for experiments to verify the effectiveness of our method. Based on simulation results, we can conclude that our method can distinguish different activity clearly.

On the Impact of Biceps Muscle Fatigue in Human Activity Recognition

Nowadays, Human Activity Recognition (HAR) systems, which use wearables and smart systems, are a part of our daily life. Despite the abundance of literature in the area, little is known about the impact of muscle fatigue on these systems’ performance. In this work, we use the biceps concentration curls exercise as an example of a HAR activity to observe the impact of fatigue impact on such systems. Our dataset consists of 3000 biceps concentration curls performed and collected from 20 volunteers aged between 20–35. Our findings indicate that fatigue often occurs in later sets of an exercise and extends the completion time of later sets by up to 31% and decreases muscular endurance by 4.1%. Another finding shows that changes in data patterns are often occurring during fatigue presence, causing seven features to become statistically insignificant. Further findings indicate that fatigue can cause a substantial decrease in performance in both subject-specific and cross-subject models. Finally, we observed that a Feedforward Neural Network (FNN) showed the best performance in both cross-subject and subject-specific models in all our evaluations.

Exploration of Human Activity Recognition Using a Single Sensor for Stroke Survivors and Able-Bodied People

Commonly used sensors like accelerometers, gyroscopes, surface electromyography sensors, etc., which provide a convenient and practical solution for human activity recognition (HAR), have gained extensive attention. However, which kind of sensor can provide adequate information in achieving a satisfactory performance, or whether the position of a single sensor would play a significant effect on the performance in HAR are sparsely studied. In this paper, a comparative study to fully investigate the performance of the aforementioned sensors for classifying four activities (walking, tooth brushing, face washing, drinking) is explored. Sensors are spatially distributed over the human body, and subjects are categorized into three groups (able-bodied people, stroke survivors, and the union of both). Performances of using accelerometer, gyroscope, sEMG, and their combination in each group are evaluated by adopting the Support Vector Machine classifier with the Leave-One-Subject-Out Cross-Validation technique, and the optimal sensor position for each kind of sensor is presented based on the accuracy. Experimental results show that using the accelerometer could obtain the best performance in each group. The highest accuracy of HAR involving stroke survivors was 95.84 ± 1.75% (mean ± standard error), achieved by the accelerometer attached to the extensor carpi ulnaris. Furthermore, taking the practical application of HAR into consideration, a novel approach to distinguish various activities of stroke survivors based on a pre-trained HAR model built on healthy subjects is proposed, the highest accuracy of which is 77.89 ± 4.81% (mean ± standard error) with the accelerometer attached to the extensor carpi ulnaris.

Improving Human Activity Recognition Performance by Data Fusion and Feature Engineering

Human activity recognition (HAR) is essential in many health-related fields. A variety of technologies based on different sensors have been developed for HAR. Among them, fusion from heterogeneous wearable sensors has been developed as it is portable, non-interventional and accurate for HAR. To be applied in real-time use with limited resources, the activity recognition system must be compact and reliable. This requirement can be achieved by feature selection (FS). By eliminating irrelevant and redundant features, the system burden is reduced with good classification performance (CP). This manuscript proposes a two-stage genetic algorithm-based feature selection algorithm with a fixed activation number (GFSFAN), which is implemented on the datasets with a variety of time, frequency and time-frequency domain features extracted from the collected raw time series of nine activities of daily living (ADL). Six classifiers are used to evaluate the effects of selected feature subsets from different FS algorithms on HAR performance. The results indicate that GFSFAN can achieve good CP with a small size. A sensor-to-segment coordinate calibration algorithm and lower-limb joint angle estimation algorithm are introduced. Experiments on the effect of the calibration and the introduction of joint angle on HAR shows that both of them can improve the CP.

Deep Neural Network for Slip Detection on Ice Surface

Slip-induced falls are among the most common causes of major occupational injuries and economic loss in Canada. Identifying the risk factors associated with slip events is key to developing preventive solutions to reduce falls. One factor is the slip-resistance quality of footwear, which is fundamental to reducing the number of falls. Measuring footwear slip resistance with the recently developed Maximum Achievable Angle (MAA) test requires a trained researcher to identify slip events in a simulated winter environment. The human capacity for information processing is limited and human error is natural, especially in a cold environment. Therefore, to remove conflicts associated with human errors, in this paper a deep three-dimensional convolutional neural network is proposed to detect the slips in real-time. The model has been trained by a new dataset that includes data from 18 different participants with various clothing, footwear, walking directions, inclined angles, and surface types. The model was evaluated on three types of slips: Maxi-slip, midi-slip, and mini-slip. This classification is based on the slip perception and recovery of the participants. The model was evaluated based on both 5-fold and Leave-One-Subject-Out (LOSO) cross validation. The best accuracy of 97% was achieved when identifying the maxi-slips. The minimum accuracy of 77% was achieved when classifying the no-slip and mini-slip trials. The overall slip detection accuracy was 86% with sensitivity and specificity of 81% and 91%, respectively. The overall accuracy dropped by about 2% in LOSO cross validation. The proposed slip detection algorithm is not only beneficial for footwear manufactures to improve their footwear slip resistance quality, but it also has other potential applications, such as improving the slip resistance properties of flooring in healthcare facilities, commercial kitchens, and oil drilling platforms.

On the Challenges and Potential of Using Barometric Sensors to Track Human Activity

Barometers are among the oldest engineered sensors. Historically, they have been primarily used either as environmental sensors to measure the atmospheric pressure for weather forecasts or as altimeters for aircrafts. With the advent of microelectromechanical system (MEMS)-based barometers and their systematic embedding in smartphones and wearable devices, a vast breadth of new applications for the use of barometers has emerged. For instance, it is now possible to use barometers in conjunction with other sensors to track and identify a wide range of human activity classes. However, the effectiveness of barometers in the growing field of human activity recognition critically hinges on our understanding of the numerous factors affecting the atmospheric pressure, as well as on the properties of the sensor itself-sensitivity, accuracy, variability, etc. This review article thoroughly details all these factors and presents a comprehensive report of the numerous studies dealing with one or more of these factors in the particular framework of human activity tracking and recognition. In addition, we specifically collected some experimental data to illustrate the effects of these factors, which we observed to be in good agreement with the findings in the literature. We conclude this review with some suggestions on some possible future uses of barometric sensors for the specific purpose of tracking human activities.

Human Activity Recognition using Inertial, Physiological and Environmental Sensors: A Comprehensive Survey

In the last decade, Human Activity Recognition (HAR) has become a vibrant research area, especially due to the spread of electronic devices such as smartphones, smartwatches and video cameras present in our daily lives. In addition, the advance of deep learning and other machine learning algorithms has allowed researchers to use HAR in various domains including sports, health and well-being applications. For example, HAR is considered as one of the most promising assistive technology tools to support elderly's daily life by monitoring their cognitive and physical function through daily activities. This survey focuses on critical role of machine learning in developing HAR applications based on inertial sensors in conjunction with physiological and environmental sensors.

A Benchmark of Data Stream Classification for Human Activity Recognition on Connected Objects

This paper evaluates data stream classifiers from the perspective of connected devices, focusing on the use case of Human Activity Recognition. We measure both the classification performance and resource consumption (runtime, memory, and power) of five usual stream classification algorithms, implemented in a consistent library, and applied to two real human activity datasets and three synthetic datasets. Regarding classification performance, the results show the overall superiority of the Hoeffding Tree, the Mondrian forest, and the Naïve Bayes classifiers over the Feedforward Neural Network and the Micro Cluster Nearest Neighbor classifiers on four datasets out of six, including the real ones. In addition, the Hoeffding Tree and—to some extent—the Micro Cluster Nearest Neighbor, are the only classifiers that can recover from a concept drift. Overall, the three leading classifiers still perform substantially worse than an offline classifier on the real datasets. Regarding resource consumption, the Hoeffding Tree and the Mondrian forest are the most memory intensive and have the longest runtime; however, no difference in power consumption is found between classifiers. We conclude that stream learning for Human Activity Recognition on connected objects is challenged by two factors which could lead to interesting future work: a high memory consumption and low F1 scores overall.

Smartphone Motion Sensor-Based Complex Human Activity Identification Using Deep Stacked Autoencoder Algorithm for Enhanced Smart Healthcare System

Human motion analysis using a smartphone-embedded accelerometer sensor provided important context for the identification of static, dynamic, and complex sequence of activities. Research in smartphone-based motion analysis are implemented for tasks, such as health status monitoring, fall detection and prevention, energy expenditure estimation, and emotion detection. However, current methods, in this regard, assume that the device is tightly attached to a pre-determined position and orientation, which might cause performance degradation in accelerometer data due to changing orientation. Therefore, it is challenging to accurately and automatically identify activity details as a result of the complexity and orientation inconsistencies of the smartphone. Furthermore, the current activity identification methods utilize conventional machine learning algorithms that are application dependent. Moreover, it is difficult to model the hierarchical and temporal dynamic nature of the current, complex, activity identification process. This paper aims to propose a deep stacked autoencoder algorithm, and orientation invariant features, for complex human activity identification. The proposed approach is made up of various stages. First, we computed the magnitude norm vector and rotation feature (pitch and roll angles) to augment the three-axis dimensions (3-D) of the accelerometer sensor. Second, we propose a deep stacked autoencoder based deep learning algorithm to automatically extract compact feature representation from the motion sensor data. The results show that the proposed integration of the deep learning algorithm, and orientation invariant features, can accurately recognize complex activity details using only smartphone accelerometer data. The proposed deep stacked autoencoder method achieved 97.13% identification accuracy compared to the conventional machine learning methods and the deep belief network algorithm. The results suggest the impact of the proposed method to improve a smartphone-based complex human activity identification framework.

Data Quality and Reliability Assessment of Wearable EMG and IMU Sensor for Construction Activity Recognition

The workforce shortage is one of the significant problems in the construction industry. To overcome the challenges due to workforce shortage, various researchers have proposed wearable sensor-based systems in the area of construction safety and health. Although sensors provide rich and detailed information, not all sensors can be used for construction applications. This study evaluates the data quality and reliability of forearm electromyography (EMG) and inertial measurement unit (IMU) of armband sensors for construction activity classification. To achieve the proposed objective, the forearm EMG and IMU data collected from eight participants while performing construction activities such as screwing, wrenching, lifting, and carrying on two different days were used to analyze the data quality and reliability for activity recognition through seven different experiments. The results of these experiments show that the armband sensor data quality is comparable to the conventional EMG and IMU sensors with excellent relative and absolute reliability between trials for all the five activities. The activity classification results were highly reliable, with minimal change in classification accuracies for both the days. Moreover, the results conclude that the combined EMG and IMU models classify activities with higher accuracies compared to individual sensor models.

Fast Wearable Sensor-Based Foot-Ground Contact Phase Classification Using a Convolutional Neural Network with Sliding-Window Label Overlapping

Classification of foot–ground contact phases, as well as the swing phase is essential in biomechanics domains where lower-limb motion analysis is required; this analysis is used for lower-limb rehabilitation, walking gait analysis and improvement, and exoskeleton motion capture. In this study, sliding-window label overlapping of time-series wearable motion data in training dataset acquisition is proposed to accurately detect foot–ground contact phases, which are composed of 3 sub-phases as well as the swing phase, at a frequency of 100 Hz with a convolutional neural network (CNN) architecture. We not only succeeded in developing a real-time CNN model for learning and obtaining a test accuracy of 99.8% or higher, but also confirmed that its validation accuracy was close to 85%.

Hand Resting Tremor Assessment of Healthy and Patients With Parkinson's Disease: An Exploratory Machine Learning Study

The aim of this study is comparing the accuracies of machine learning algorithms to classify data concerning healthy subjects and patients with Parkinson’s Disease (PD), toward different time window lengths and a number of features. Thirty-two healthy subjects and eighteen patients with PD took part on this study. The study obtained inertial recordings by using an accelerometer and a gyroscope assessing both hands of the subjects during hand resting state. We extracted time and temporal frequency domain features to feed seven machine learning algorithms: k-nearest-neighbors (kNN); logistic regression; support vector classifier (SVC); linear discriminant analysis; random forest; decision tree; and gaussian Naïve Bayes. The accuracy of the classifiers was compared using different numbers of extracted features (i.e., 272, 190, 136, 82, and 27) from different time window lengths (i.e., 1, 5, 10, and 15 s). The inertial recordings were characterized by oscillatory waveforms that, especially in patients with PD, peaked in a frequency range between 3 and 8 Hz. Outcomes showed that the most important features were the mean frequency, linear prediction coefficients, power ratio, power density skew, and kurtosis. We observed that accuracies calculated in the testing phase were higher than in the training phase. Comparing the testing accuracies, we found significant interactions among time window length and the type of classifier (p < 0.05). The study found significant effects on estimated accuracies, according to their type of algorithm, time window length, and their interaction. kNN presented the highest accuracy, while SVC showed the worst results. kNN feeding by features extracted from 1 and 5 s were the combination with more frequently highest accuracies. Classification using few features led to similar decision of the algorithms. Moreover, performance increased significantly according to the number of features used, reaching a plateau around 136. Finally, the results of this study suggested that kNN was the best algorithm to classify hand resting tremor in patients with PD.

Wearable Inertial Sensors for Daily Activity Analysis Based on Adam Optimization and the Maximum Entropy Markov Model

Advancements in wearable sensors technologies provide prominent effects in the daily life activities of humans. These wearable sensors are gaining more awareness in healthcare for the elderly to ensure their independent living and to improve their comfort. In this paper, we present a human activity recognition model that acquires signal data from motion node sensors including inertial sensors, i.e., gyroscopes and accelerometers. First, the inertial data is processed via multiple filters such as Savitzky–Golay, median and hampel filters to examine lower/upper cutoff frequency behaviors. Second, it extracts a multifused model for statistical, wavelet and binary features to maximize the occurrence of optimal feature values. Then, adaptive moment estimation (Adam) and AdaDelta are introduced in a feature optimization phase to adopt learning rate patterns. These optimized patterns are further processed by the maximum entropy Markov model (MEMM) for empirical expectation and highest entropy, which measure signal variances for outperformed accuracy results. Our model was experimentally evaluated on University of Southern California Human Activity Dataset (USC-HAD) as a benchmark dataset and on an Intelligent Mediasporting behavior (IMSB), which is a new self-annotated sports dataset. For evaluation, we used the “leave-one-out” cross validation scheme and the results outperformed existing well-known statistical state-of-the-art methods by achieving an improved recognition accuracy of 91.25%, 93.66% and 90.91% when compared with USC-HAD, IMSB, and Mhealth datasets, respectively. The proposed system should be applicable to man–machine interface domains, such as health exercises, robot learning, interactive games and pattern-based surveillance.

Modelling Patient Behaviour Using IoT Sensor Data: a Case Study to Evaluate Techniques for Modelling Domestic Behaviour in Recovery from Total Hip Replacement Surgery

The UK health service sees around 160,000 total hip or knee replacements every year and this number is expected to rise with an ageing population. Expectations of surgical outcomes are changing alongside demographic trends, whilst aftercare may be fractured as a result of resource limitations. Conventional assessments of health outcomes must evolve to keep up with these changing trends. Health outcomes may be assessed largely by self-report using Patient Reported Outcome Measures (PROMs), such as the Oxford Hip or Oxford Knee Score, in the months up to and following surgery. Though widely used, many PROMs have methodological limitations and there is debate about how to interpret results and definitions of clinically meaningful change. With the development of a home-monitoring system, there is opportunity to characterise the relationship between PROMs and behaviour in a natural setting and to develop methods of passive monitoring of outcome and recovery after surgery. In this paper, we discuss the motivation and technology used in long-term continuous observation of movement, sleep and domestic routine for healthcare applications, such as the HEmiSPHERE project for hip and knee replacement patients. In this case study, we evaluate trends evident in data of two patients, collected over a 3-month observation period post-surgery, by comparison with scores from PROMs for sleep and movement quality, and by comparison with a third control home. We find that accelerometer and indoor localisation data correctly highlight long-term trends in sleep and movement quality and can be used to predict sleep and wake times and measure sleep and wake routine variance over time, whilst indoor localisation provides context for the domestic routine and mobility of the patient. Finally, we discuss a visual method of sharing findings with healthcare professionals.

The new design of cows' behavior classifier based on acceleration data and proposed feature set

Monitor and classify behavioral activities in cows is a helpful support solution for livestock based on the analysis of data from sensors attached to the animal. Accelerometers are particularly suited for monitoring cow behaviors due to small size, lightweight and high accuracy. Nevertheless, the interpretation of the data collected by such sensors when characterizing the type of behaviors still brings major challenges to developers, related to activity complexity (i.e., certain behaviors contain similar gestures). This paper presents a new design of cows' behavior classifier based on acceleration data and proposed feature set. Analysis of cow acceleration data is used to extract features for classification using machine learning algorithms. We found that with 5 features (mean, standard deviation, root mean square, median, range) and 16-second window of data (1 sample/second), classification of seven cow behaviors (including feeding, lying, standing, lying down, standing up, normal walking, active walking) achieved the overall highest performance. We validated the results with acceleration data from a public source. Performance of our proposed classifier was evaluated and compared to existing ones in terms of the sensitivity, the accuracy, the positive predictive value, and the negative predictive value.

A Model-Based Design Floating-Point Accumulator. Case of Study: FPGA Implementation of a Support Vector Machine Kernel Function

Recent research in wearable sensors have led to the development of an advanced platform capable of embedding complex algorithms such as machine learning algorithms, which are known to usually be resource-demanding. To address the need for high computational power, one solution is to design custom hardware platforms dedicated to the specific application by exploiting, for example, Field Programmable Gate Array (FPGA). Recently, model-based techniques and automatic code generation have been introduced in FPGA design. In this paper, a new model-based floating-point accumulation circuit is presented. The architecture is based on the state-of-the-art delayed buffering algorithm. This circuit was conceived to be exploited in order to compute the kernel function of a support vector machine. The implementation of the proposed model was carried out in Simulink, and simulation results showed that it had better performance in terms of speed and occupied area when compared to other solutions. To better evaluate its figure, a practical case of a polynomial kernel function was considered. Simulink and VHDL post-implementation timing simulations and measurements on FPGA confirmed the good results of the stand-alone accumulator.

Human Behavior Recognition Based on Motion Data Analysis

The development of sensor technologies and smart devices has made it possible to realize real-time data acquisition of human beings. Human behavior monitoring is the process of obtaining activity information with wearables and computer technology. In this paper, we design a data preprocessing method based on the data collected by a single three-axis accelerometer. We first use Butterworth filter as low-pass filtering to remove the noise. Then, we propose a KGA algorithm to remove abnormal data and smooth them at the same time. This method uses genetic algorithm to optimize the parameters of Kalman filter. After that, we use a threshold-based method to identify falls that are harmful to the elderly. The key point of this method is to distinguish falls from people’s daily activities. According to the characteristics of human falls, we extract eigenvalues that can effectively distinguish daily activities from falls. In addition, we use cross-validation to determine the threshold of the method. The results show that in the analysis of 11 kinds of human daily activities and 15 types of falls, our method can distinguish 15 types of falls. The recognition recall rate in our method reaches 99.1%.