6 August 2025 in Biomedical Signal Processing and Control
27 October 2021 in IEEE Transactions on Vehicular Technology
In the existing millimeter-wave (mmWave) wireless positioning systems, the method based on sparse Bayesian learning (SBL) uses the channel sparsity to estimate the parameters required for positioning, such as angle of arrival (AOA) and time delay, but most existing SBL solutions only consider the angle sparsity. In this paper, we consider the joint sparsity of the angle domain and time delay domain to propose an improved SBL algorithm by using a new two-dimensional adaptive grid refinement method in the SBL framework. This algorithm solves the grid mismatch problem caused by the fixed grid in the traditional SBL method, and reduces the algorithm complexity of the off-grid SBL (OGSBL) algorithm. We also obtain the Cramér-Rao bound (CRB) of AOA, time delay and position estimation based on the mmWave multipath signals to analyze the estimation errors. Simulation results show that the performance of the proposed algorithm is better than existing algorithms and can approach CRB.
21 November 2022 in IEEE Access
The etiology of Temporomandibular disorders (TMD) is still unclear, and its symptoms, signs and progression are extremely complex. TMD require early diagnosis and treatments, especially for combinations with other oral diseases. The research targets at developing an artificial neural network (ANN) model for predicting TMD based on clinical-collected data including clinical features, systematic medical condition, and psychosocial state. The popular data mining-based ANN was utilized to predict TMD with all 18 variables collected from patients as the input. The total dataset consists of 88 cases which were reviewed by Board-certificated orthodontists. 75% (66) cases are randomly selected as the training dataset, while the remaining 25% (22) cases are for test. Among the considered 88 cases, 58 (65.9%) were with TMD, while the left 30 (34.1%) without TMD. The numbers of male and female were 21 and 67, respectively, while the average age was 27.63 years. The calculated average sensitivity and specificity of ANN-based TMD risk predictions through 10-fold-cross-validation analysis were 92.31% (95% confidence interval (CI), 62.09%-99.60%) and 88.89% (95% CI, 50.67%-99.42%), respectively. Moreover, the accuracy rate of ANN was 90.91% (95% CI, 78.90%-100.00%). The results show the proposed ANN model could predict the TMD risks with a high accuracy rate, indicating the potential of machine learning in oral and maxillofacial diseases screening and diagnosis, which was further illustrated in a comparison with two doctors. This study can help dental care providers to find individuals’ risk of TMD by inputting patient’s psychological factors, oral examinations, and systemic medical conditions to the developed artificial intelligence (AI) model.
27 February 2025 in IEEE Internet of Things Journal
With the rapid growth of Internet of Medical Things (IoMT) devices and the advancement of medical Large Language Models (LLMs), smart healthcare applications including regular vital sign monitoring, medical intervention, and medication adherence tracking, play an increasingly important role in the near future. As the data generated by smart healthcare applications usually have strict requirements for latency and privacy, traditional data processing in a centralized manner is not the preferred choice, especially for users in remote areas. In this paper, we consider the Non-Terrestrial Networks (NTNs) composed of Low Earth Orbit (LEO) satellites and Unmanned Aerial Vehicles (UAVs) to realize seamless coverage. A federated reinforcement learning based hierarchical computation offloading strategy is proposed to make distinct decisions and protect data privacy. Moreover, we optimize the communication efficiency between IoMT devices, UAVs, and satellites, improve task processing capabilities, and reduce network load by reasonably setting task priorities. The simulation results show that the proposed method performs well in reducing energy consumption and shortening task completion time, significantly improving resource utilization and achieving system load balancing.
1 January 2026 in IEEE Open Journal of the Computer Society
Adenoid hypertrophy is a common pediatric and adolescent condition that can cause mouth breathing, snoring, adenoidal facies, and craniofacial developmental issues. In clinical ENT diagnosis, medical imaging is often required to evaluate adenoid status. In this paper, we introduce a lightweight YOLO-based model for automatic detection and segmentation of adenoids and upper airway regions in lateral cephalometric radiographs. The model incorporates lightweight depthwise separable convolutions and an improved attention fusion module (C3K2CSA), significantly reducing parameter size and computation cost. Experimental results show a Box mAP$_{50}$ of 0.922 and a Mask mAP$_{50}$ of 0.94, with reductions of 17.5% in model size and 13.7% in computation. These results demonstrate the model’s effectiveness and suitability for clinical deployment.