Eddy current testing (ECT) has been employed as a traditional non-destructive testing and evaluation (NDT&E) tool for many years. It has developed from single frequency to multiple frequencies, and eventually to pulsed and swept-frequency excitation. Recent progression of wireless power transfer (WPT) and flexible printed devices open opportunities to address challenges of defect detection and reconstruction under complex geometric situations. In this paper, a transmitter-receiver (Tx-Rx) flexible printed coil (FPC) array that uses the WPT approach featuring dual resonance responses for the first time has been proposed. The dual resonance responses can provide multiple parameters of samples, such as defect characteristics, lift-offs and material properties, while the flexible coil array allows area mapping of complex structures. To validate the proposed approach, experimental investigations of a single excitation coil with multiple receiving coils using the WPT principle were conducted on a curved pipe surface with a natural dent defect. The FPC array has one single excitation coil and 16 receiving (Rx) coils, which are used to measure the dent by using 21 C-scan points on the dedicated dent sample. The experimental data were then used for training and evaluation of dual resonance responses in terms of multiple feature extraction, selection and fusion for quantitative NDE. Four features, which include resonant magnitudes and principal components of the two resonant areas, were investigated for mapping and reconstructing the defective dent through correlation analysis for feature selection and feature fusion by deep learning. It shows that deep learning-based multiple feature fusion has outstanding performance for 3D defect reconstruction of WPT-based FPC-ECT. This article is part of the theme issue 'Advanced electromagnetic non-destructive evaluation and smart monitoring'.
Since various dance teaching systems have attracted much attention with the development of Artificial Intelligence (AI) technology, this paper improves the recognition performance of Latin dance teaching systems by optimizing the action recognition model. Firstly, the object detection and action recognition technology under the current AI technology is analyzed, and the Two-stage object detection algorithm and One-stage object detection algorithm are evaluated. Secondly, the technologies and functions contained in the adolescent health Latin dance teaching system are described, including image acquisition, feature extraction, object detection, and action recognition. Finally, the action recognition algorithm is optimized based on object detection, and the rationality and feasibility of the proposed algorithm are verified by experiments. The experimental results show that the optimization algorithm can search the optimal feature subset after five iterations on Undefine Classes of 101 (UCF101) dataset, but it needs seven iterations on Human Motion Database 51 (HMDB51) dataset. Meanwhile, when using support vector machine classifier, the optimization algorithm can achieve the highest accuracy of motion recognition. Regressive Function, Multinomial Naive Bayes and Gaussian Naive Bayes Algorithms have lower prediction delay, as low as 0.01s. Therefore, this paper has certain reference significance for the design and implementation of adolescent health Latin dance teaching system.
Diabetic retinopathy (DR) is the leading cause of vision loss among older adults. The goal of this case-control study was to identify circulating miRNAs for the diagnosis of DR. The miRNeasy Serum/Plasma Kit was used to extract serum miRNAs. The μParaflo™ MicroRNA microarray was used to detect the expression levels of the miRNAs. The miRWalk algorithm was applied to predict the target genes of the miRNAs, which were further confirmed by the dual luciferase reporter gene system in HEK293T cells. A microarray was performed between 5 DR cases and 5 age-, sex-, body mass index-, and duration of diabetes-matched type 2 diabetic (T2DM) controls. The quantitative reverse transcription polymerase chain reaction technique was used to validate the differentially expressed circulating miRNAs in 45 DR cases and 45 well-matched controls. Receiver operating characteristic (ROC) curve analysis was used to evaluate the performance of the circulating miRNAs as diagnostic biomarkers for DR. Our microarray analysis screened out miR-2116-5p and miR-3197 as significantly up-regulated in DR cases compared with the controls. Furthermore, two miRNAs were validated in the 45 DR cases and 45 controls. The ROC analysis suggested that both miR-3197 and miR-2116-5p distinguished DR cases from controls. An additional dual-luciferase reporter gene assay confirmed that notch homolog 2 (NOTCH2) was the target gene of miR-2116-5p. Both miR-3197 and miR-2116-5p were identified as promising diagnostic biomarkers for DR. Future research is still needed to explore the molecular mechanisms of miR-3197 and miR-2116-5p in the pathogenesis of DR.