Search

Neurological diseases are a significant health threat, often presenting through abnormalities in electroencephalogram (EEG) signals during seizures. In recent years, machine learning (ML) technologies have been explored as a means of automated EEG pathology diagnosis. However, existing ML-based EEG binary classification methods largely focus on extracting EEG-related features, which may lead to poor performance in classifying EEG signals by overlooking potentially redundant information. In this paper, we propose a novel Kruskal–Wallis (KW) test-based framework for EEG pathology detection. Our framework first divides EEG data into frequency sub-bands using wavelet packet decomposition and then extracts statistical characteristics from each selected coefficient. Next, the piecewise aggregation approximation technique is used to obtain the aggregated feature vectors, followed by the KW statistical test methodology to select significant features. Finally, three ensemble learning classifiers, random forest, categorical boosting (CatBoost), and light gradient boosting machine, are used to classify the extracted significant features into normal or abnormal classes. Our proposed framework achieves an accuracy of 89.13%, F1-score of 87.60%, and G-mean of 88.60%, respectively, outperforming other competing techniques on the same dataset, which shows the great promise in EEG pathology detection. [ABSTRACT FROM AUTHOR]

In this study, a Glutinous rice 89-1 (Gr 89-1) × Shuhui 527 recombinant inbred line population (RIL) comprising 309 F9-generations was used to screen gradient molecular markers. The phenotypic variation and distribution of eight agronomic traits obtained from multiyear and multilocation samples, as well as the network expression relationships between agronomic traits and molecular markers, were investigated. The results showed that there were 14 phenotypic lines with significant differences in the RILs, and the molecular testing results of most of the lines were consistent with the phenotype. The correlation degree between the first-level molecular markers and the eight agronomic traits was 100%. Excluding the correlations of third-level markers with grain width and grain length, the degree of correlation between molecular markers and agronomic traits decreased with an increase in marker levels. The RILs were divided into eight core populations and one approximate population, revealing genetic correspondence between agronomic traits and molecular markers. [ABSTRACT FROM AUTHOR]

The innovative Incremental Sheet Forming (ISF) process affects the post-forming properties of thermoplastic polymers. However, the effects of degree of plastic strain, and the orientation and size of specimen on the mechanical properties are still unknown. In the present study, therefore, the ISF process is performed on a polymer sheet by varying the plastic strain ranging from 6% to 108%. The corresponding effects on the properties and associated polymer structure are quantified by conducting a variety of mechanical and structural tests. The results reveal that the post-ISF tensile properties like yield stress, ultimate stress, drawing stress, elastic modulus and elongation decrease from 26.6 to 10 MPa, 30.5 to 15.4 MPa, 18.9 to 9.9 MPa, 916 to 300 MPa and 1107% to 457%, respectively, as the strain increases in the investigated range. The value of post-ISF relaxation properties, contrary to the tensile properties, increases with increasing strain up to 62%. Particularly, reductions in stress, strain and modulus increase from 41% to 202%, 37% to 51%, and 41% to 202%. As regard the orientation effect, the sheet in the feed direction shows greater strength than the transverse direction (up to 142% in yield stress and 72% in ultimate stress). Moreover, the smaller sample offers greater strength than the larger one (up to 158% in yield stress and 109% in ultimate stress). The analysis of the post-ISF tensile properties and structural results lead us to conclude that the drop in the tensile properties due to increasing strain occurs due to corresponding increase in the voids area fraction (1.25% to 31%) and a reduction in the crystallinity (38% to 31%). [ABSTRACT FROM AUTHOR]

In order to improve the initialization robustness of visual inertial SLAM, the complementarity of the optical flow method and the feature-based method can be used in vision data processing. The parallel initialization method is proposed, where the optical flow inertial initialization and the monocular feature-based initialization are carried out at the same time. After the initializations, the state estimation results are jointly optimized by bundle adjustment. The proposed method retains more mapping information, and correspondingly is more adaptable to the initialization scene. It is found that the initialization map constructed by the proposed method features a comparable accuracy to the one constructed by ORB-SLAM3 in monocular inertial mode. Since the online extrinsic parameter estimation can be realized by the proposed method, it is considered better than ORB-SLAM3 in the aspect of portability. By the experiments performed on the benchmark dataset EuRoC, the effectiveness and robustness of the proposed method are validated. [ABSTRACT FROM AUTHOR]