Hybrid-Pattern Recognition Modeling with Arrhythmia Signal Processing for Ubiquitous Health Management.

We established a web-based ubiquitous health management (UHM) system, "ECG4UHM", for processing ECG signals with AI-enabled models to recognize hybrid arrhythmia patterns, including atrial premature atrial complex (APC), atrial fibrillation (AFib), ventricular premature complex (VPC), and ventricular tachycardia (VT), versus normal sinus rhythm (NSR). The analytical model coupled machine learning methods, such as multiple layer perceptron (MLP), random forest (RF), support vector machine (SVM), and naive Bayes (NB), to process the hybrid patterns of four arrhythmia symptoms for AI computation. The data pre-processing used Hilbert-Huang transform (HHT) with empirical mode decomposition to calculate ECGs' intrinsic mode functions (IMFs). The area centroids of the IMFs' marginal Hilbert spectrum were suggested as the HHT-based features. We engaged the MATLABTM compiler and runtime server in the ECG4UHM to build the recognition modules for driving AI computation to identify the arrhythmia symptoms. The modeling extracted the crucial data sets from the MIT-BIH arrhythmia open database. The validated models, including the premature pattern (i.e., APC-VPC) and the fibril-rapid pattern (i.e., AFib-VT) against NSR, could reach the best area under the curve (AUC) of the receiver operating characteristic (ROC) of approximately 0.99. The models for all hybrid patterns, without VPC versus AFib and VT, achieved an average accuracy of approximately 90%. With the prediction test, the respective AUCs of the NSR and APC versus the AFib, VPC, and VT were 0.94 and 0.93 for the RF and SVM on average. The average accuracy and the AUC of the MLP, RF, and SVM models for APC-VT reached the value of 0.98. The self-developed system with AI computation modeling can be the backend of the intelligent social-health system that can recognize hybrid arrhythmia patterns in the UHM and home-isolated cares.

AIoT-Enabled Rehabilitation Recognition System-Exemplified by Hybrid Lower-Limb Exercises.

Ubiquitous health management (UHM) is vital in the aging society. The UHM services with artificial intelligence of things (AIoT) can assist home-isolated healthcare in tracking rehabilitation exercises for clinical diagnosis. This study combined a personalized rehabilitation recognition (PRR) system with the AIoT for the UHM of lower-limb rehabilitation exercises. The three-tier infrastructure integrated the recognition pattern bank with the sensor, network, and application layers. The wearable sensor collected and uploaded the rehab data to the network layer for AI-based modeling, including the data preprocessing, featuring, machine learning (ML), and evaluation, to build the recognition pattern. We employed the SVM and ANFIS methods in the ML process to evaluate 63 features in the time and frequency domains for multiclass recognition. The Hilbert-Huang transform (HHT) process was applied to derive the frequency-domain features. As a result, the patterns combining the time- and frequency-domain features, such as relative motion angles in y- and z-axis, and the HHT-based frequency and energy, could achieve successful recognition. Finally, the suggestive patterns stored in the AIoT-PRR system enabled the ML models for intelligent computation. The PRR system can incorporate the proposed modeling with the UHM service to track the rehabilitation program in the future.

Personalized Rehabilitation Recognition for Ubiquitous Healthcare Measurements.

The physical therapeutic application needs personalized rehabilitation recognition (PRR) for ubiquitous healthcare measurements (UHMs). This study employed the adaptive neuro-fuzzy inference system (ANFIS) to generate a PRR model for a self-development system of UHM. The subjects wore a sensor-enabled wristband during physiotherapy exercises to measure the scheduled motions of their limbs. In the model, the sampling data collected from the scheduled motions are labeled by an arbitrary number within a defined range. The sample datasets are referred as the design of an initial fuzzy inference system (FIS) with data preprocessing, feature visualizing, fuzzification, and fuzzy logic rules. The ANFIS then processes data training to adjust the FIS for optimization. The trained FIS then can infer the motion labels via defuzzification to recognize the features in the test data. The average recognition rate was higher than 90% for the testing motions if the subject followed the sampling schedule. With model implementation, the middle section of motion datasets in each second is recommended for recognition in the UHM system which also includes a mobile App to retrieve the personalized FIS in order to trace the exercise. This approach contributes a PRR model with trackable diagrams for the physicians to explore the rehabilitation motions in details.

Developing a ubiquitous health management system with healthy diet control for metabolic syndrome healthcare in Taiwan.

BACKGROUND AND OBJECTIVE: Self-management in healthcare can allow patients managing their health data anytime and everywhere for prevention of chronic diseases. This study established a prototype of ubiquitous health management system (UHMS) with healthy diet control (HDC) for people who need services of metabolic syndrome healthcare in Taiwan. METHODS: System infrastructure comprises of three portals and a database tier with mutually supportive components to achieve functionality of diet diaries, nutrition guides, and health risk assessments for self-health management. With the diet, nutrition, and personal health database, the design enables the analytical diagrams on the interactive interface to support a mobile application for diet diary, a Web-based platform for health management, and the modules of research and development for medical care. For database integrity, dietary data can be stored at offline mode prior to transformation between mobile device and server site at online mode. RESULTS: The UHMS-HDC was developed by open source technology for ubiquitous health management with personalized dietary criteria. The system integrates mobile, internet, and electronic healthcare services with the diet diary functions to manage healthy diet behaviors of users. The virtual patients were involved to simulate the self-health management procedure. The assessment functions were approved by capturing the screen snapshots in the procedure. The proposed system development was capable for practical intervention. CONCLUSION: This approach details the expandable framework with collaborative components regarding the self-developed UHMS-HDC. The multi-disciplinary applications for self-health management can support the healthcare professionals to reduce medical resources and improve healthcare effects for the patient who requires monitoring personal health condition with diet control. The proposed system can be practiced for intervention in the hospital.

Fuzzy Computing Model of Activity Recognition on WSN Movement Data for Ubiquitous Healthcare Measurement.

Ubiquitous health care (UHC) is beneficial for patients to ensure they complete therapeutic exercises by self-management at home. We designed a fuzzy computing model that enables recognizing assigned movements in UHC with privacy. The movements are measured by the self-developed body motion sensor, which combines both accelerometer and gyroscope chips to make an inertial sensing node compliant with a wireless sensor network (WSN). The fuzzy logic process was studied to calculate the sensor signals that would entail necessary features of static postures and dynamic motions. Combinations of the features were studied and the proper feature sets were chosen with compatible fuzzy rules. Then, a fuzzy inference system (FIS) can be generated to recognize the assigned movements based on the rules. We thus implemented both fuzzy and adaptive neuro-fuzzy inference systems in the model to distinguish static and dynamic movements. The proposed model can effectively reach the recognition scope of the assigned activity. Furthermore, two exercises of upper-limb flexion in physical therapy were applied for the model in which the recognition rate can stand for the passing rate of the assigned motions. Finally, a web-based interface was developed to help remotely measure movement in physical therapy for UHC.

An activity recognition model using inertial sensor nodes in a wireless sensor network for frozen shoulder rehabilitation exercises.

This paper proposes a model for recognizing motions performed during rehabilitation exercises for frozen shoulder conditions. The model consists of wearable wireless sensor network (WSN) inertial sensor nodes, which were developed for this study, and enables the ubiquitous measurement of bodily motions. The model employs the back propagation neural network (BPNN) algorithm to compute motion data that are formed in the WSN packets; herein, six types of rehabilitation exercises were recognized. The packets sent by each node are converted into six components of acceleration and angular velocity according to three axes. Motor features such as basic acceleration, angular velocity, and derivative tilt angle were input into the training procedure of the BPNN algorithm. In measurements of thirteen volunteers, the accelerations and included angles of nodes were adopted from possible features to demonstrate the procedure. Five exercises involving simple swinging and stretching movements were recognized with an accuracy of 85%-95%; however, the accuracy with which exercises entailing spiral rotations were recognized approximately 60%. Thus, a characteristic space and enveloped spectrum improving derivative features were suggested to enable identifying customized parameters. Finally, a real-time monitoring interface was developed for practical implementation. The proposed model can be applied in ubiquitous healthcare self-management to recognize rehabilitation exercises.

Poly-silicon nanowire sensor for sodium chloride concentration measurement.

In this paper, the poly-Si nanowire sensor was fabricated by top-down technique for sodium chloride concentration measurement. The results showed that the smallest threshold voltage and the best resolution were 1.65 V and 0.41 µM, respectively. Furthermore, the sensor can be reused more than 50 times which maintained acceptable performance and showed good linearity of the calibration within wide range of the concentration. Based on these results, it can be concluded that the proposed sensor has great potential to be used for measuring complicated sample with suitable modification on the surface of nanowires.

The backend design of an environmental monitoring system upon real-time prediction of groundwater level fluctuation under the hillslope.

The groundwater level represents a critical factor to evaluate hillside landslides. A monitoring system upon the real-time prediction platform with online analytical functions is important to forecast the groundwater level due to instantaneously monitored data when the heavy precipitation raises the groundwater level under the hillslope and causes instability. This study is to design the backend of an environmental monitoring system with efficient algorithms for machine learning and knowledge bank for the groundwater level fluctuation prediction. A Web-based platform upon the model-view controller-based architecture is established with technology of Web services and engineering data warehouse to support online analytical process and feedback risk assessment parameters for real-time prediction. The proposed system incorporates models of hydrological computation, machine learning, Web services, and online prediction to satisfy varieties of risk assessment requirements and approaches of hazard prevention. The rainfall data monitored from the potential landslide area at Lu-Shan, Nantou and Li-Shan, Taichung, in Taiwan, are applied to examine the system design.