Convolutional neural networks (CNNs) have shown great promise in human activity recognition, but long-term dependencies in time series data can be difficult to capture using standard CNNs. This study introduces a new CNN architecture that incorporates a multi-head attention mechanism (CNN-MHA) to address this challenge. This mechanism is composed of several attention heads, each independently calculating attention weights for distinct segments of the input. The attention head outputs are then concatenated and processed through a fully connected layer to produce the final attention representation. The multi-head attention mechanism allows the network to focus on relevant features and maintain long-term dependencies in the input data. The proposed model is evaluated on the physical activity monitoring for aging people dataset (PAMAP2) from the UCI machine learning repository, which is preprocessed by cleaning, normalization, segmentation, and reshaping before splitting into training, validation, and testing sets. The experimental results demonstrate that the CNN-MHA model outperforms existing models, achieving F1-score of 95.7%. Particularly, the multi-head attention mechanism significantly improves the model’s ability to recognize complex activity patterns. Furthermore, our model attained an average inference latency of 0.304 seconds, which can be crucial in real-time applications. The findings clearly demonstrate the substantial promise of the proposed CNN-MHA architecture for optimizing human activity recognition tasks, offering a powerful tool for advancing the state-of-the-art in this domain.

As the Platform for Advanced Wireless Research (PAWR) in rural broadband, the ARA wireless living lab features the deployment of first-of-its-kind wireless access and backhaul platforms in real-world agriculture and rural settings, and preliminary experiments have demonstrated very promising results, e.g., up to 3.2 Gbps wireless access throughput and more than 10 Gbps throughput across a wireless backhaul link of over 10 km. ARA is expected to be publicly released for broad community use starting in September 2023. Through this demo, we plan to share, for the first time, with the wireless research community the transformative research experiments enabled by ARA. To stimulate discussion and community participation, we will demonstrate a few example experiments ranging from MU-MIMO in TV White Space (TVWS) bands to long-range mmWave and microwave backhaul communications, as well as open-source 5G NR protocol stacks such as srsRAN and OpenAirInterface.

The status quo of limited broadband connectivity in rural areas motivates the need for fielding alternatives such as long-distance wireless mesh networks. A key aspect of fielding wireless meshes cost-effectively is planning how to connect the last-mile networks to the core network service providers (i.e., the network between the edge access terminals and the landline / optical fiber terminals) with minimal infrastructure cost and throughput constraints. This so-called middle-mile network optimization, which includes topology construction, tower height assignment, antenna and orientation selection, as well as transmit power assignment, is known to be a computationally hard problem. In this paper, we provide the first polynomial time approximation solution for a generalized version of the middle-mile network optimization problem, wherein point-to-point (i.e., WiFi p2p) links are deployed to bridge last-mile networks. Our solution has a cost performance ratio of O(ln|A| + |B|/|A| + |A|+|B|/γ), where A and B respectively denote the number of terminals and nonterminals and γ is the ratio of link capacity/terminal demand. Furthermore, our solution extends to hybrid networks, i.e., point-to-multipoint (i.e., WiFi p2mp) or omnidirectional (i.e., TV White Space) can serve as hyperlinks in addition to point-to-point links, to further reduce the cost of wireless links. We provide a complementary heuristic for our middle-mile network optimization solution that adds hyperlinks if and only if they reduce the cost.

In recent years, the number of yoga practitioners has been drastically increased and there are more men and older people practice yoga than ever before. Internet of Things (IoT)-based yoga training system is needed for those who want to practice yoga at home. Some studies have proposed RGB/Kinect camera-based or wearable device-based yoga posture recognition methods with a high accuracy; however, the former has a privacy issue and the latter is impractical in the long-term application. Thus, this paper proposes an IoT-based privacy-preserving yoga posture recognition system employing a deep convolutional neural network (DCNN) and a low-resolution infrared sensor-based wireless sensor network (WSN). The WSN has three nodes (x, y, and z-axes) where each integrates 8 × 8 pixels' thermal sensor module and a Wi-Fi module for connecting the deep learning server. We invited 18 volunteers to perform 26 yoga postures for two sessions each lasted for 20 s. First, recorded sessions are saved as .csv files, then preprocessed and converted to grayscale posture images. Totally, 93200 posture images are employed for the validation of the proposed DCNN models. The tenfold cross-validation results revealed that F1-scores of the models trained with xyz (all 3-axes) and y (only y-axis) posture images were 0.9989 and 0.9854, respectively. An average latency for a single posture image classification on the server was 107 ms. Thus, we conclude that the proposed IoT-based yoga posture recognition system has a great potential in the privacy-preserving yoga training system.

Erectile dysfunction (ED) affects millions of men worldwide. Men with ED generally complain failure to attain or maintain an adequate erection during sexual activity. The prevalence of ED is strongly correlated with age, affecting about 40% of men at age 40 and nearly 70% at age 70. A variety of chronic diseases, including diabetes, ischemic heart disease, congestive heart failure, hypertension, depression, chronic renal failure, obstructive sleep apnea, prostate disease, gout, and sleep disorder, were reported to be associated with ED. In this study, data retrieved from a subset of the National Health Insurance Research Database of Taiwan were used for designing the clinical decision support system (CDSS) for predicting ED incidences in men. The positive cases were male patients aged 20-65 who were diagnosed with ED between January 2000 and December 2010 confirmed by at least three outpatient visits or at least one inpatient visit, while the negative cases were randomly selected from the database without a history of ED and were frequency (1:1), age, and index year matched with the ED patients. Data of a total of 2832 ED patients and 2832 non-ED patients, each consisting of 41 features including index age, 10 comorbidities, and 30 other comorbidity-related variables, were retrieved for designing the predictive models. Integrated genetic algorithm and support vector machine was adopted to design the CDSSs with two experiments of independent training and testing (ITT) conducted to verify their effectiveness. In the 1st ITT experiment, data extracted from January 2000 till December 2005 (61.51%, 1742 positive cases and 1742 negative cases) were used for training and validating and the data retrieved from January 2006 till December 2010 were used for testing (38.49%), whereas in the 2nd ITT experiment, data in the training set (77.78%) were extracted from January 2000 till Deceber 2007 and those in the testing set (22.22%) were retrieved afterward. Tenfold cross validation and three different objective functions were adopted for obtaining the optimal models with best predictive performance in the training phase. The testing results show that the CDSSs achieved a predictive performance with accuracy, sensitivity, specificity, g-mean, and area under ROC curve of 74.72%-76.65%, 72.33%-83.76%, 69.54%-77.10%, 0.7468-0.7632, and 0.766-0.817, respectively. In conclusion, the CDSSs designed based on cost-sensitive objective functions as well as salient comorbidity-related features achieve satisfactory predictive performance for predicting ED incidences.

Patient readmission may be attributed to unsuccessful treatments, new diseases, or worsening comorbidities, which manifest degraded healthcare quality and increased cost of a hospital. Clinical decision support systems (CDSSs) provide useful information and expert knowledge to improve diagnostic performance, treatment outcomes, and healthcare quality in clinical settings. The objectives of this study is to design a CDSS based on the data retrieved from the National Health Insurance Research Database (NHIRD) for assisting physicians in the identification of high-risk patients of 30-day readmissions. A wrapper method integrating genetic algorithm (GA) and support vector machine (SVM) was used to designing the CDSSs by selecting salient features and designing the classifiers based on 2 under-sampling methods and 3 objective functions. The accuracy, sensitivity, specificity, and area under ROC curve (AUC) of the CDSSs achieve 69.33-71.44%, 66.27-69.41%, 69.32-72.24%, and 0.7518-0.7601, respectively, which outperform the models designed based on data retrieved from electronic medical records. An online web service system is provided to assist physicians for detecting potential pneumonia patients who have higher probability of all-cause readmission after being discharged from hospitals, thereby giving them necessary interventions to reduce morbidities and mortalities and to reduce healthcare cost.

In this study, the resource blocks (RB) are allocated to user equipment (UE) according to the evolutional algorithms for long term evolution (LTE) systems. Particle Swarm Optimization (PSO) algorithm is one of the evolutionary algorithms, based on the imitation of a flock of birds foraging behavior through learning and grouping the best experience. Therefore, we propose a Simple Particle Swarm Optimization (SPSO) algorithm for RB allocation to enhance the throughput of Device-to-Device (D2D) communications and improve the system capacity performance. The simulation results show that with less population size of M=10, the SPSO can perform quickly convergence to sub-optimal solution in the 100'hgeneration. Therefore, as compared to the random allocation (Rand) method, the proposed SPSO can obtain sub-optimum performance with more 2 UEs than the Rand method.

Inhaled corticosteroid is a popular treatment for chronic respiratory diseases. Its side effects include decreased bone mineral density and osteoporosis. The aim of this study is to investigate the association of inhaled corticosteroids and fracture. The data of the patients who were 20 years old or older visiting healthcare centers (outpatients or inpatients) and had been prescribed with inhale corticosteroids within 2002–2010 were retrieved from the National Health Insurance Research Database (NHIRD) for this investigation. Patients diagnosed with hip fracture or vertebrate fractures before using inhaled corticosteroid were excluded. A Total of 11645 patients receiving inhaled corticosteroid therapy were included in this study; among them, 1134 (9.7%) with either hip fracture or vertebrate fracture were observed. The average cumulated inhaled corticosteroid for patients with fracture (230.14±442.76 mg) was significantly higher (p<0.001) than those without fracture (171.97±318.13 mg). Based on the analytical results, it was observed that long-term use of inhaled corticosteroids may induce osteoporosis and exhibit higher incidence of hip or vertebrate fractures.

The goal of this research is to develop a wearable electrocardiogram (ECG) apparatus with Bluetooth 4.0 function which can detect the heart beats, mental index and physical activity in real time. It measured one lead ECG signal. Its specification follows the standard of International Electrotechnical Commission (IEC) for medical electrical equipment and ambulatory electrocardiographic systems (Holter ECG system). Moreover, its ingress protection rating has IP 68 degree. Thus, the user could wear it all day and to do any exercise including swimming. In order to let patients feeling comfortable, its weight is light and it could full run above 24 hours. The data was stored in a 4G flash memory. We also proposed a method to detect the heart beats and frequency parameters of heart rate variability (HRV) in real time. This algorithm can be embedded in a single chip microcontroller. We used the MIT-BIH Arrhythmia Database to evaluate the performance of our designed system. The sensitivity and positive predictivity arrives 99.7% and 99.0%. The results show our designed Bluetooth wearable ECG apparatus can record the ECG signal and detect the heart beats, HRV and step counts in real time. It is very suitable to be applied in the ubiquitous healthcare.

The ability to track joint positions for Microsoft Kinect is useful for rehabilitation in both clinical and home settings. Unfortunately, most available commercial Kinect video games are not developed for rehabilitation purposes. In this study, a low-cost rehabilitation game training system has been developed for patients with movement disorders, it is suitable for use at home under the distant supervision of a therapist. This research explores the potential and the limitations of the Kinect-based system in the applications of e-rehabilitation. A prototypic system including 4 video games was developed for e-rehabilitation. In order to test the usefulness and effectiveness of the Kinect Virtual Rehabilitation System (KVRS), the Balance Recovery game was used to train patients with stroke to recover their balance abilities. It was demonstrated that balance training using the Kinect e-rehabilitation system exhibited a significant effect in recovering balance function for the stroke patients. Additionally, the outcome of questionnaire survey revealed that the KVRS video game was perceived as effective and easy in operation.

Cytologic screening has been widely used for detecting the cervical cancers. In this study, a semiautomatic PC-based cellular image analysis system was developed for segmenting nuclear and cytoplasmic contours and for computing morphometric and textual features to train support vector machine (SVM) classifiers to classify four different types of cells and to discriminate dysplastic from normal cells. A software program incorporating function, including image reviewing and standardized denomination of file names, was also designed to facilitate and standardize the workflow of cell analyses. Two experiments were conducted to verify the classification performance. The cross-validation results of the first experiment showed that average accuracies of 97.16% and 98.83%, respectively, for differentiating four different types of cells and in discriminating dysplastic from normal cells have been achieved using salient features (8 for four-cluster and 7 for two-cluster classifiers) selected with SVM recursive feature addition. In the second experiment, 70% (837) of the cell images were used for training and 30% (361) for testing, achieving an accuracy of 96.12% and 98.61% for four-cluster and two-cluster classifiers, respectively. The proposed system provides a feasible and effective tool in evaluating cytologic specimens.

Due to hardware design or cost consideration, sensors may possess sector-like sensing coverage. Furthermore, by stepper motors, sensors can rotate to cover the objects around them. This type of sensors are called rotatable and directional (R&D) sensors. Through rotation, R&D sensors provide temporal coverage to objects by "periodically” detecting their existence. In the paper, we first develop an event-driven surveillance system by R&D sensors, where objects are monitored by the sensors equipped with infrared detectors and cameras. When an object is taken away, the sensor monitoring the object reports a warning message along with detailed snapshots from the surroundings. Then, motivated by the system, we formulate an R&D sensor deployment problem, which tries to deploy the minimum number of R&D sensors to cover a given set of objects such that each object is covered by 0 < \delta \le 1 ratio of time in every frame. We show this problem to be NP-hard and propose two efficient heuristics. The maximum covering deployment (MCD) heuristic iteratively deploys a sensor to cover more objects, and performs well when objects congregate together. The disk-overlapping deployment (DOD) heuristic deploys sensors to cover the joint sectors of overlapped disks, so it works better when objects are arbitrarily placed in the sensing field. The paper contributes in defining a new temporal coverage model by R&D sensors, developing a surveillance application for this model, and proposing efficient heuristics to reduce the deployment cost.

Traditional restaurant service is typically passive: Waiters must interact with customers directly before processing their orders. However, a high-quality service system should be customer centered; it should immediately recognize customer identities, favorite menus, and expenditure records to provide customer-centric services. To achieve this goal, this study integrates radio frequency identification (RFID), wireless local area network, database technologies, and a menu recommender to develop an intelligent e-restaurant for customer-centric service. This system enables waiters to immediately identify customers via RFID-based membership cards and then actively recommend the most appropriate menus through menu recommender for customers. Experimental results that are obtained from a case study conducted in two Taipei restaurants indicate that the proposed system has practical potential in providing customer-centric service.

Ventilator weaning is the process of discontinuing mechanical ventilation from patients with respiratory failure. Ventilator support should be withdrawn as soon as possible when it is no longer necessary in order to reduce the likelihood of known nosocomial complications and costs. Previous investigation indicated that clinicians were often wrong when predicting weaning outcome. The motivation of this study is that although successful ventilator weaning of ICU patients has been widely studied, indicators for accurate prediction are still under investigation. The goal of this study is to find a prediction model for successful ventilator weaning using variables such physiological variables, clinical syndromes, demographic variables, and other useful information. The data obtained from 231 patients who had been supported by mechanical ventilator for longer than 21 days within the period from Nov. 2002 to Dec. 2005 were studied retrospectively. Among them, 188 patients were recruited from the period within Nov. 2002 to Dec. 2004 and the other 43 patients from Jan. 2004 to Dec. 2005. All the patients were clinically stable before being considered to undergo a weaning trial. Twenty-seven variables in total were collected with only 6 variables reaching significant level (p<0.05) were used for support vector machine (SVM) classification after statistical analysis. The results show that the constructed model is valuable in assisting clinical doctors to decide if a patient is ready to wean from the ventilator with the sensitivity, specificity, and accuracy as high as 94.74%, 95.83%, and 95.35%, respectively. Further prospective bed side test is needed to verify the efficacy of the model.