Multi-Scale Attention-Based Deep Neural Network for Brain Disease Diagnosis

Whole brain functional connectivity(FC)patterns obtained from resting-state functional magnetic resonance imaging(rs-fMRI)have been widely used in the diagnosis of brain disorders such as autism spectrum disorder(ASD).Recently,an increasing number of studies have focused on employing deep learning techniques to analyze FC patterns for brain disease classification.However,the high dimensionality of the FC features and the interpretation of deep learning results are issues that need to be addressed in the FC-based brain disease classification.In this paper,we proposed a multi-scale attention-based deep neural network(MSA-DNN)model to classify FC patterns for the ASD diagnosis.The model was implemented by adding a flexible multi-scale attention(MSA)module to the auto-encoder based backbone DNN,which can extract multi-scale features of the FC patterns and change the level of attention for different FCs by continuous learning.Our model will reinforce the weights of important FC features while suppress the unimportant FCs to ensure the sparsity of the model weights and enhance the model interpretability.We performed systematic experiments on the large multi-sites ASD dataset with both ten-fold and leaveone-site-out cross-validations.Results showed that our model outperformed classical methods in brain disease classification and revealed robust intersite prediction performance.We also localized important FC features and brain regions associated with ASD classification.Overall,our study further promotes the biomarker detection and computer-aided classification for ASD diagnosis,and the proposed MSA module is flexible and easy to implement in other classification networks.

HIGHER ORDER SPECTRAL ANALYSIS IN FAULT DIAGNOSIS OF ROTORS

The nonlinear properties of rotating machinery vibration signals are presented. The relationship between faults and quadratic phase coupling is discussed. The mechanism that gives rise to quadratic phase coupling is analyzed, and the coupling models are summarized. As a result, higher order spectra analysis is introduced into fault diagnosis of rotors. A brief review of the properties of higher order spectra is presented. Furthermore, the bicoherence spectrum is employed to extract the features that signify the machinery condition. Experiments show that bicoherence spectrum patterns of different faults are quite different, so it is proposed to identify the faults in rotors.

Fault Feature Extraction of Rolling Bearing Based on an Improved Cyclical Spectrum Density Method

The traditional cyclical spectrum density（CSD） method is widely used to analyze the fault signals of rolling bearing. All modulation frequencies are demodulated in the cyclic frequency spectrum. Consequently, recognizing bearing fault type is difficult. Therefore, a new CSD method based on kurtosis（CSDK） is proposed. The kurtosis value of each cyclic frequency is used to measure the modulation capability of cyclic frequency. When the kurtosis value is large, the modulation capability is strong. Thus, the kurtosis value is regarded as the weight coefficient to accumulate all cyclic frequencies to extract fault features. Compared with the traditional method, CSDK can reduce the interference of harmonic frequency in fault frequency, which makes fault characteristics distinct from background noise. To validate the effectiveness of the method, experiments are performed on the simulation signal, the fault signal of the bearing outer race in the test bed, and the signal gathered from the bearing of the blast furnace belt cylinder. Experimental results show that the CSDK is better than the resonance demodulation method and the CSD in extracting fault features and recognizing degradation trends. The proposed method provides a new solution to fault diagnosis in bearings.

Sex Differences in Diagnosis and Clinical Phenotypes of Chinese Children with Autism Spectrum Disorder

The aim of this study was to explore the differences between boys and girls in the diagnosis and clinical phenotypes of autism spectrum disorder（ASD） in China＇s mainland. Children diagnosed with ASD（n = 1064, 228 females） were retrospectively included in the analysis. All children were assessed using the Autism Diagnostic Interview-Revised（ADI-R） and Autism Diagnostic Observation Schedule（ADOS）. The results showed that girls scored significantly higher in ADI-R socioemotional reciprocity than boys, and also scored lower in ADI-R and ADOS restricted and repetitive behaviors（RRBs）. Meanwhile, the proportions of girls who satisfied the diagnostic cut-off scores in the ADI-R RRBs domain were lower than in boys（P ／ 0.05）. Our results indicated that girls with ASD show greater socio-emotional reciprocity than boys. Girls also tended to show fewer RRBs than boys, and the type of RRBs in girls differ from those in boys. The ADI-R was found to be less sensitive in girls, particularly for assessment in the RRBs domain.

An Overview of Autism Spectrum Disorder, Heterogeneity and Treatment Options

Since the documented observations of Kanner in1943, there has been great debate about the diagnoses, the sub-types, and the diagnostic threshold that relates to what is now known as autism spectrum disorder（ASD）.Re？ecting this complicated history, there has been continual re？nement from DSM-III with ‘Infantile Autism＇ to the current DSM-V diagnosis. The disorder is now widely accepted as a complex, pervasive, heterogeneous condition with multiple etiologies, sub-types, and developmental trajectories. Diagnosis remains based on observation of atypical behaviors, with criteria of persistent de？cits in social communication and restricted and repetitive patterns of behavior. This review provides a broad overview of the history, prevalence, etiology, clinical presentation, and heterogeneity of ASD. Factors contributing to heterogeneity, including genetic variability, comorbidity, and gender are reviewed. We then explore current evidencebased pharmacological and behavioral treatments for ASD and highlight the complexities of conducting clinical trials that evaluate therapeutic ef？cacy in ASD populations.Finally, we discuss the potential of a new wave of research examining objective biomarkers to facilitate the evaluation of sub-typing, diagnosis, and treatment response in ASD.