Search

Purpose: Single-shot (SS) EPI is widely used for clinical DWI. This study aims to develop an end-to-end deep learning-based method with a novel loss function in an improved network structure to simultaneously increase the resolution and correct distortions for SS-EPI DWI., Theory and Methods: Point-spread-function (PSF)-encoded EPI can provide high-resolution, distortion-free DWI images. A distorted image from SS-EPI can be described as the convolution between a PSF function with a distortion-free image. The deconvolution process to recover the distortion-free image can be achieved with a convolution neural network, which also learns the mapping function between low-resolution SS-EPI and high-resolution reference PSF-EPI to achieve superresolution. To suppress the oversmoothing effect, we proposed a modified generative adversarial network structure, in which a dense net with gradient map guidance and a multilevel fusion block was used as the generator. A fractional anisotropy loss was proposed to utilize the diffusion anisotropy information among diffusion directions. In vivo brain DWI data were used to test the proposed method., Results: The results show that distortion-corrected high-resolution DWI images with restored structural details can be obtained from low-resolution SS-EPI images by taking advantage of the high-resolution anatomical images. Additionally, the proposed network can improve the quantitative accuracy of diffusion metrics compared with previously reported networks., Conclusion: Using high-resolution, distortion-free EPI-DWI images as references, a deep learning-based method to simultaneously increase the perceived resolution and correct distortions for low-resolution SS-EPI was proposed. The results show that DWI image quality and diffusion metrics can be improved., (© 2023 International Society for Magnetic Resonance in Medicine.)

In recent years, deep learning‐based fault diagnosis technology for high‐voltage circuit breakers (HVCB) has advanced significantly, but the working environment of HVCBs is complex, resulting in unsatisfactory fault diagnosis results of HVCBs in noisy environment and existing deep learning methods are difficult to solve this problem. This paper proposes a multi‐channel convolutional neural network combines dense residual structure and attention mechanism to achieve high‐precision and high‐robust diagnosis of HVCBs in noisy backgrounds. A dense residual network is introduced into the convolutional neural network to prevent feature loss during network propagation to preserve the difference information between the network layers as much as possible, Simultaneously, a channel attention mechanism is introduced to adaptively adjust the weights of different convolution channels. The model can extract multi‐scale features from the original signal and fully exploit the intrinsic relationship between the vibration signal and the HVCB's operating state. The experimental results show that the diagnostic method can still meet the requirements of fault diagnosis in the presence of noise, with an average diagnostic accuracy rate of 85.92% when the signal‐to‐noise ratio is −4. The model outperforms the traditional single‐channel model in terms of diagnostic accuracy and stability. [ABSTRACT FROM AUTHOR]

COVID-19 (coronavirus disease 2019) is a new viral infection disease that is widely spread worldwide. Deep learning plays an important role in COVID-19 images diagnosis. This paper reviews the recent progress of deep learning in COVID-19 images applications from five aspects; Firstly, 33 COVID-19 datasets and data enhancement methods are introduced; Secondly, COVID-19 classification methods based on supervised learning are summarized from four aspects of VGG, ResNet, DenseNet and Lightweight Networks. The COVID-19 segmentation methods based on supervised learning are summarized from four aspects of attention mechanism, multiscale mechanism, residual connectivity mechanism, and dense connectivity mechanism; Thirdly, the application of deep learning in semi-supervised COVID-19 images diagnosis in terms of consistency regularization methods and self-training methods. Fourthly, the application of deep learning in unsupervised COVID-19 diagnosis in terms of autoencoder methods and unsupervised generative adversarial methods. Moreover, the challenges and future work of COVID-19 images diagnostic methods in the field of deep learning are summarized. This paper reviews the latest research status of COVID-19 images diagnosis in deep learning, which is of positive significance to the detection of COVID-19. [ABSTRACT FROM AUTHOR]

Deep learning (DL) has attracted significant attention in partial discharge (PD) pattern recognition of gas‐insulated switchgear (GIS). However, the existing DL method extracts numerical information and ignores spatial correlation information. In addition, once the sample size is reduced, model performance is severely degraded. To address these issues, a novel capsule deep graph convolutional network (CDGCN) is employed for GIS PD pattern recognition. First, the GIS PD data is transformed to graph data. Then, the CDGCN is established for GIS PD pattern recognition. The introduction of a ChannelSortPooling layer enriches the graph representation information and filters out the feature values with higher information content. In addition, the multi‐view graph representation is converted to a capsule form, and a capsule network with stronger feature extraction capability is introduced to improve the accuracy of the final graph representation. The results reveal that the proposed CDGCN can effectively diagnose the GIS PD and is more accurate than the traditional DL method. The proposed CDGCN can handle small samples pattern recognition problems and provides a reliable reference for GIS PD pattern recognition. [ABSTRACT FROM AUTHOR]

Dense convolutional network (DenseNet) is a hot topic in deep learning research in recent years, which has good applications in medical image analysis. In this paper, DenseNet is summarized from the following aspects. First, the basic principle of DenseNet is introduced; second, the development of DenseNet is summarized and analyzed from five aspects: broaden DenseNet structure, lightweight DenseNet structure, dense unit, dense connection mode, and attention mechanism; finally, the application research of DenseNet in the field of medical image analysis is summarized from three aspects: pattern recognition, image segmentation, and object detection. The network structures of DenseNet are systematically summarized in this paper, which has certain positive significance for the research and development of DenseNet. [ABSTRACT FROM AUTHOR]

Purpose: This study aims to propose a novel algorithm for slab boundary artifact correction in both single‐band multislab imaging and simultaneous multislab (SMSlab) imaging. Theory and Methods: In image domain, the formation of slab boundary artifacts can be regarded as modulating the artifact‐free images using the slab profiles and introducing aliasing along the slice direction. Slab boundary artifact correction is the inverse problem of this process. An iterative algorithm based on convolutional neural networks (CNNs) is proposed to solve the problem, termed CNN‐enabled inversion for slab profile encoding (CPEN). Diffusion‐weighted SMSlab images and reference images without slab boundary artifacts were acquired in 7 healthy subjects for training. Images of 5 healthy subjects were acquired for testing, including single‐band multislab and SMSlab images with 1.3‐mm or 1‐mm isotropic resolution. CNN‐enabled inversion for slab profile encoding was compared with a previously reported method (i.e., nonlinear inversion for slab profile encoding [NPEN]). Results: CNN‐enabled inversion for slab profile encoding reduces the slab boundary artifacts in both single‐band multislab and SMSlab images. It also suppresses the slab boundary artifacts in the diffusion metric maps. Compared with NPEN, CPEN shows fewer residual artifacts in different acquisition protocols and more significant improvements in quantitative assessment, and it also accelerates the computation by more than 35 times. Conclusion: CNN‐enabled inversion for slab profile encoding can reduce the slab boundary artifacts in multislab acquisitions. It shows better slab boundary artifact correction capacity, higher robustness, and computation efficiency when compared with NPEN. It has the potential to improve the accuracy of multislab acquisitions in high‐resolution DWI and functional MRI. [ABSTRACT FROM AUTHOR]

• 1D-ACCN model is used to optimize HVCB fault identification. • Attention mechanism and capsule network enhances the feature extraction capability. • The dynamic routing algorithm is employed for capsule construction and training. • The results demonstrate its superiority over the other traditional methods. The traditional fault diagnosis method of high-voltage circuit breakers (HVCBs) based on deep learning largely depends on the huge data size. When the data sample size is insufficient, it is difficult to meet the needs of diagnosis performance. To improve the diagnostic performance, a one-dimensional attention convolutional capsule neural network is proposed in this paper. First, to take advantage of the time fine-grained information of the vibration signals, a one-dimensional convolutional neural network with a wide convolution kernel is designed, thus avoiding the tedious manual feature extraction of the original data. Then, an attention mechanism is added in the pooling layer to change the weight allocation of vibration signal fragments and increase its feature extraction capability. Finally, a capsule layer is introduced between the convolution layer and the fully connected layer to represent feature vectorization, which ensures the integrity of fault feature extraction. The classification structure digital capsule is obtained by the feature transfer of dynamic routing. The experiment results show that the combination of the attention mechanism and capsule layer is feasible and the accuracy can reach 94.7% when the number of samples is small. Compared with the traditional method, the proposed method still has high diagnostic accuracy and anti-noise ability in the case of small sample. [ABSTRACT FROM AUTHOR]

• A novel UN-CN is proposed for few-shot HVCB fault diagnosis. • U-net is introduced to extract discriminative features with insufficient samples. • Capsule network is used to reduce feature loss and solve the few-shot problem. • The dynamic routing algorithm is employed for model optimization. • The results demonstrate the superiority of UN-CN in few-shot fault diagnosis. Deep learning methods have achieved noteworthy seeing results in the mechanical fault diagnosis of high-voltage circuit breakers with the recent advancements in artificial intelligence. However, the premise of the above method for obtaining excellent performance is to have sufficient samples, which is impractical due to the characteristics of the high-voltage circuit breakers. This study proposes a novel U-Net with CapsNet for high-voltage circuit breakers fault diagnosis to resolve these issues, achieving a high-precision and robust diagnosis of few-shot high-voltage circuit breakers. In a few-shot diagnosis, the U-Net with CapsNet takes advantage of the high accuracy of the U-Net. The capsule network is used in the contraction and expansion paths of the original U-Net to reduce the loss of features in the pooling process. The forward transfer from the bottom to the high-level capsule is completed by the dynamic routing algorithm. The feature information in the high-level capsule is unified with the bottom layer. The experimental results show that using the U-Net with CapsNet proposal, we can quickly and accurately realize the fault diagnosis of few-shot high-voltage circuit breakers, with an accuracy of 93.25%. The model has faster convergence speed and better stability, which provides a reliable solution for efficient and accurate fault diagnosis of high-voltage circuit breakers compared with traditional methods. [ABSTRACT FROM AUTHOR]