Positional multi-length and mutual-attention network for epileptic seizure classification.

The automatic classification of epilepsy electroencephalogram (EEG) signals plays a crucial role in diagnosing neurological diseases. Although promising results have been achieved by deep learning methods in this task, capturing the minute abnormal characteristics, contextual information, and long dependencies of EEG signals remains a challenge. To address this challenge, a positional multi-length and mutual-attention (PMM) network is proposed for the automatic classification of epilepsy EEG signals. The PMM network incorporates a positional feature encoding process that extracts minute abnormal characteristics from the EEG signal and utilizes a multi-length feature learning process with a hierarchy residual dilated LSTM (RDLSTM) to capture long contextual dependencies. Furthermore, a mutual-attention feature reinforcement process is employed to learn the global and relative feature dependencies and enhance the discriminative abilities of the network. To validate the effectiveness PMM network, we conduct extensive experiments on the public dataset and the experimental results demonstrate the superior performance of the PMM network compared to state-of-the-art methods.

Research on predicting the diffusion of toxic heavy gas sulfur dioxide by applying a hybrid deep learning model to real case data.

Toxic heavy gas sulfur dioxide (SO2) is a specific life and environmental hazard. Predicting the diffusion of SO2 has become a research focus in fields such as environmental and safety studies. However, traditional methods, such as kinetic models, cannot balance precision and time. Thus, they do not meet the needs of emergency decision-making. Deep learning (DL) models are emerging as a highly regarded solution, providing faster and more accurate predictions of gas concentrations. To this end, this study proposes an innovative hybrid DL model, the parallel-connected convolutional neural network-gated recurrent unit (PC CNN-GRU). This model utilizes two CNNs connected in parallel to process gas release and meteorological datasets, enabling the automatic extraction of high-dimensional data features and handling of long-term temporal dependencies through the GRU. The proposed model demonstrates good performance (RMSE, MAE, and R2 of 20.1658, 10.9158, and 0.9288, respectively) with real data from the Project Prairie Grass (PPG) case. Meanwhile, to address the issue of limited availability of raw data, in this study, time series generative adversarial network (TimeGAN) are introduced for SO2 diffusion studies for the first time, and their effectiveness is verified. To enhance the practicality of the research, the contribution of drivers to SO2 diffusion is quantified through the utilization of the permutation importance (PIMP) and Sobol' method. Additionally, the maximum safe distance downwind under various conditions is visualized based on the SO2 toxicity endpoint concentration. The results of the analyses can provide a scientific basis for relevant decisions and measures.

Intelligent identification of natural gas pipeline defects based on improved pollination algorithm.

As a natural gas pipeline approaches the end of its service life, the integrity of the pipeline starts failing because of corrosion or cracks. These and other defects affect the normal production and operation of the pipeline. Therefore, the identification of pipeline defects is critical to ensure the normal, safe, and efficient operation of these pipelines. In this study, a combination of adaptive adjustment based on conversion probability and Gaussian mutation strategy was used to improve the flower pollination algorithm (FPA) and enhance the search ability of traditional flower pollination. The adaptive adjustment of the transition probability effectively balances the development and exploration abilities of the algorithm. The improved flower pollination algorithm (IFPA) outperformed six classical benchmark functions that were used to verify the superiority of the improved algorithm. A Gaussian mutation strategy was integrated with IFPA to optimise the initial input weights and thresholds of the extreme learning machine (ELM), improve the balance and exploration ability of the algorithm, and increase the efficiency and accuracy for identifying pipeline defects. The proposed IFPA-ELM model for pipeline defect identification effectively overcomes the tendency of FPA to converge to local optima and that of ELM to engage in overfitting, which cause poor recognition accuracy. The identification rates of various pipeline defects by the IFPA-ELM algorithm are 97% and 96%, which are 34% and 13% higher, respectively, than those of FPA and FPA-ELM. The IFPA-ELM model may be used in the intelligent diagnosis of pipeline defects to solve practical engineering problems. Additionally, IFPA could be further optimised with respect to the time dimension, parameter settings, and general adaptation for application to complex engineering optimisation problems in various fields.

Investigation of the Solubility of Elemental Sulfur (S) in Sulfur-Containing Natural Gas with Machine Learning Methods.

Some natural gases are toxic because they contain hydrogen sulfide (H2S). The solubility pattern of elemental sulfur (S) in toxic natural gas needs to be studied for environmental protection and life safety. Some methods (e.g., experiments) may pose safety risks. Measuring sulfur solubility using a machine learning (ML) method is fast and accurate. Considering the limited experimental data on sulfur solubility, this study used consensus nested cross-validation (cnCV) to obtain more information. The global search capability and learning efficiency of random forest (RF) and weighted least squares support vector machine (WLSSVM) models were enhanced via a whale optimization-genetic algorithm (WOA-GA). Hence, the WOA-GA-RF and WOA-GA-WLSSVM models were developed to accurately predict the solubility of sulfur and reveal its variation pattern. WOA-GA-RF outperformed six other similar models (e.g., RF model) and six other published studies (e.g., the model designed by Roberts et al.). Using the generic positional oligomer importance matrix (gPOIM), this study visualized the contribution of variables affecting sulfur solubility. The results show that temperature, pressure, and H2S content all have positive effects on sulfur solubility. Sulfur solubility significantly increases when the H2S content exceeds 10%, and other conditions (temperature, pressure) remain the same.

Cross-Modal Prostate Cancer Segmentation via Self-Attention Distillation.

The automatic and accurate segmentation of the prostate cancer from the multi-modal magnetic resonance images is of prime importance for the disease assessment and follow-up treatment plan. However, how to use the multi-modal image features more efficiently is still a challenging problem in the field of medical image segmentation. In this paper, we develop a cross-modal self-attention distillation network by fully exploiting the encoded information of the intermediate layers from different modalities, and the generated attention maps of different modalities enable the model to transfer significant and discriminative information that contains more details. Moreover, a novel spatial correlated feature fusion module is further employed for learning more complementary correlation and non-linear information of different modality images. We evaluate our model in five-fold cross-validation on 358 MRI images with biopsy confirmed. Without bells and whistles, our proposed network achieves state-of-the-art performance on extensive experiments.

Barriers and facilitators to physical activity among ethnic Chinese children: a qualitative systematic review.

OBJECTIVE: The review aimed to synthesize the barriers and facilitators from the available studies that explored physical activity among ethnic Chinese children and uncover any differences or similarities in these barriers and facilitators. INTRODUCTION: Physical activity promotes overall health, fitness, and well-being in children, yet prevalence of this has been low among ethnic Chinese children who reside in either Chinese and non-Chinese territories. Research has been conducted to explore the barriers and facilitators to physical activity among ethnic Chinese children. However, no qualitative systematic review has been conducted to synthesize these barriers and facilitators. INCLUSION CRITERIA: Studies were considered for inclusion if they explored the barriers and facilitators to physical activity among ethnic Chinese children aged six to 17 years in either Chinese or non-Chinese territories, or among people who had responsibility for them in school, home, and community settings. The review included studies that focused on their views, experiences, attitudes, understandings, perceptions, and perspectives. Studies were included if they focused on qualitative data including, but not limited to, designs such as phenomenology, ethnography, grounded theory and action research. In addition, the authors considered cross-sectional surveys to find any free text relating to the review question. METHODS: MEDLINE, Embase, CINAHL, PsycINFO, BNI, AMED, Web of Science, Scopus, CNKI, Wanfang and VIP databases were searched to identify published studies. The search for unpublished studies included EThOS, OpenGrey, ProQuest Dissertations and Theses, CNKI and Wanfang. Databases were searched from their inception dates to 10 December 2018 and no language restrictions were applied. The JBI guidelines for qualitative systematic reviews were followed in conducting the review. The JBI process of meta-aggregation was used to identify categories and synthesize findings. RESULTS: Out of 9460 records identified, 11 qualitative studies met the eligibility criteria and were included in the review. Using the JBI checklist for qualitative research (10 criteria), the critical appraisal scores of the majority of studies ranged from a moderate score of 6 (n = 1) to a high score of 7 and above (n = 9). Seven studies were from China, two from Australia, one each from the United Kingdom and the United States. The sample size ranged from 12 to 115 participants. A total of 56 findings were extracted and aggregated into 21 categories, based on the similarity of meaning. From studies conducted in the Chinese territories, four synthesized findings (personal, socio-cultural, environmental, and policy- and program-related barriers and facilitators) were aggregated from 37 extracted findings and 14 aggregated categories. From studies conducted in the non-Chinese territories, only two synthesized findings (personal and socio-cultural barriers and facilitators) were derived from 19 extracted findings and seven aggregated categories. Based on the ConQual scores, confidence in the synthesized findings was moderate. CONCLUSIONS: Four broad themes emerged from the participants' accounts, namely personal, socio-cultural, environmental, and policy- and program-related factors. Barriers and facilitators at the personal and socio-cultural level (e.g., parents and teachers) were most frequently cited, reflecting the importance of children's self-influence and the role of adults. Future interventions are needed to address the identified barriers and enhance the facilitators. SYSTEMATIC REVIEW REGISTRATION NUMBER: PROSPERO CRD42018097124.

A Dynamic Multi-Scale Network for EEG Signal Classification.

Accurate and automatic classification of the speech imagery electroencephalography (EEG) signals from a Brain-Computer Interface (BCI) system is highly demanded in clinical diagnosis. The key factor in designing an automatic classification system is to extract essential features from the original input; though many methods have achieved great success in this domain, they may fail to process the multi-scale representations from different receptive fields and thus hinder the model from achieving a higher performance. To address this challenge, in this paper, we propose a novel dynamic multi-scale network to achieve the EEG signal classification. The whole classification network is based on ResNet, and the input signal first encodes the features by the Short-time Fourier Transform (STFT); then, to further improve the multi-scale feature extraction ability, we incorporate a dynamic multi-scale (DMS) layer, which allows the network to learn multi-scale features from different receptive fields at a more granular level. To validate the effectiveness of our designed network, we conduct extensive experiments on public dataset III of BCI competition II, and the experimental results demonstrate that our proposed dynamic multi-scale network could achieve promising classification performance in this task.