Search

Wei Xu,1,* Zikai Zhang,2,* Kerong Hu,1,* Ping Fang,1 Ran Li,1 Dehong Kong,1 Miao Xuan,1 Yang Yue,3 Dunmin She,4,5 Ying Xue1 1Department of Endocrinology and Metabolism, Tongji Hospital, School of Medicine, Tongji University, Shanghai, People’s Republic of China; 2Department of Oncology, Tongji Hospital, School of Medicine, Tongji University, Shanghai, People’s Republic of China; 3School of Mathematics and Statistics, University of Melbourne, Melbourne, Australia; 4Clinical Medical College, Yangzhou University, Yangzhou, Jiangsu, People’s Republic of China; 5Department of Endocrinology, Northern Jiangsu People’s Hospital Affiliated to Yangzhou University, Yangzhou, Jiangsu, People’s Republic of China*These authors contributed equally to this workCorrespondence: Ying Xue, Department of Endocrinology and Metabolism, Tongji Hospital, School of Medicine, Tongji University, No. 389, Xincun Road, Shanghai, People’s Republic of China, Tel +86-21-66111061, Email doxy0622@163.com Dunmin She, Clinical Medical College, Yangzhou University, No. 98, Nantong West Road, Yangzhou, Jiangsu, People’s Republic of China, Email sdm1979@126.comPurpose: The objective of this study was to employ machine learning (ML) models utilizing non-invasive factors to achieve early and low-cost identification of MetS in a large physical examination population.Patients and Methods: The study enrolled 9171 participants who underwent physical examinations at Northern Jiangsu People’s Hospital in 2009 and 2019, to determine MetS based on criteria established by the Chinese Diabetes Society. Non-invasive characteristics such as gender, age, body mass index (BMI), systolic blood pressure (SBP), and diastolic blood pressure (DBP) were collected and used as input variables to train and evaluate ML models for MetS identification. Several ML models were used for MetS identification, including logistic regression (LR), k-nearest neighbors algorithm (k-NN), naive bayesian (NB), decision tree (DT), random forest (RF), artificial neural network (ANN), and support vector machine (SVM).Results: Our ML models all showed good performance in the 10-fold cross-validation except for the SVM model. In the external validation, the NB model exhibited the best performance with an AUC of 0.976, accuracy of 0.923, sensitivity of 98.32%, and specificity of 91.32%.Conclusion: This study proposed a new non-invasive method for early and low-cost identification of MetS by using ML models. This approach has the potential to serve as a highly sensitive, convenient, and cost-effective tool for large-scale MetS screening.Keywords: metabolic syndrome, machine learning methods, non-invasive method, naive Bayesian

IntroductionIn the elder population, both low hemoglobin (Hb)/anemia and osteoporosis (OP) are highly prevalent. However, the relationship between Hb and OP is still poorly understood. This study was to evaluate the correlation between Hb and OP in Chinese elderly population.MethodsOne thousand and sisty-eight individuals aged 55–85 years were enrolled into this cross-sectional study during June 2019–November 2019. Data on the demographics and clinical characteristics were recorded. Detections of complete blood count, liver/kidney function, glucose metabolism and lipid profile, and thoracolumbar X-ray were performed, and bone mineral density (BMD) at lumbar spine 1–4, femur neck, and total hip was measured by dual-energy X-ray absorptiometry (DXA). Univariate and multivariate linear regression analyses were employed to evaluate the correlation between Hb with BMD T-score. Logistic regression analysis was performed to access the correlation between different Hb levels and the odds ratio (OR) for OP.ResultsCompared with non-OP group, OP patients had lower level of Hb. Univariate linear regression analysis indicated Hb level was positively related to the BMD of lumbar spine 1–4, femur neck and total hip, and this relationship remained after adjusting confounding variables [gender, age, body mass index (BMI), diabetes mellitus (DM) and morphological vertebral fracture]. Logistic regression analysis showed the ORs for OP decreased with the increase of Hb. Compared with the subjects with the lowest quartile of Hb, the OR for OP in the highest quartile group was 0.60 (0.41–0.89) after adjusting for gender, age and BMI, and the OR for OP was 0.62 (0.41–0.92) after further adjustment for gender, age, BMI, DM, and lipid indexes.DiscussionIn conclusion, Lower Hb level is related to lower BMD in the elderly population. However, whether Hb level could be used to predict the risk of OP needs to be further determined in more longitudinal clinical studies.

Mobile Edge Computing (MEC) as an emerging paradigm utilizing cloudlet or fog nodes to extend remote cloud computing to the edge of the network, is foreseen as a key technology towards next generation wireless networks. By offloading computation intensive tasks from resource constrained mobile devices to fog nodes or the remote cloud, the energy of mobile devices can be saved and the computation capability can be enhanced. For fog nodes, they can rent the resource rich remote cloud to help them process incoming tasks from mobile devices. In this architecture, the benefit of short computation and computation delay of mobile devices can be fully exploited. However, existing studies mostly assume fog nodes possess unlimited computing capacity, which is not practical, especially when fog nodes are also energy constrained mobile devices. To provide incentive of fog nodes and reduce the computation cost of mobile devices, we provide a cost effective offloading scheme in mobile edge computing with the cooperation between fog nodes and the remote cloud with task dependency constraint. The mobile devices have limited budget and have to determine which task should be computed locally or sent to the fog. To address this issue, we first formulate the offloading problem as a task finish time inimization problem with given budgets of mobile devices, which is NP-hard. We then devise two more algorithms to study the network performance. Simulation results show that the proposed greedy algorithm can achieve the near optimal performance. On average, the Brute Force method and the greedy algorithm outperform the simulated annealing algorithm by about 28.13% on the application finish time., Comment: 10 pages, 9 figures, Technical Report

Due to the high bandwidth of ultra-wideband (UWB) signals, channel estimation is regarded as a challenging issue involved in the realization of digital UWB receivers. Fortunately, the compressed sensing (CS) theory provides a promising way to solve this problem with much lower sampling rate. In traditional CS-based channel estimation schemes, researchers estimate the whole channel waveform at a time. Usually, the implementation complexity is nontrivial. In this paper, we propose a segmented CS-based channel estimation scheme. Specifically, the channel waveform is estimated in a segment by segment manner. For each segment, both the sparsity of the signal and the required scale of the measurement matrix are relatively lower. As a result, the implementation complexity can be reduced effectively. In order to improve the channel estimation performance in noisy environments, making use of the prior information on the UWB signals, we propose a weighted orthogonal matching pursuit (WOMP) algorithm. In WOMP, a weighted matching progress is performed such that dominating components of the target signal can be recovered more reliably. Simulation results show that good performance can be achieved by using our proposed scheme.