Search

Elimination of intra-artifacts in EEG has been overlooked in most of the existing sleep staging systems, especially in deep learning-based approaches. Whether intra-artifacts, originated from the eye movement, chin muscle firing, or heart beating, etc., in EEG signals would lead to a positive or a negative masking effect on deep learning-based sleep staging systems was investigated in this paper. We systematically analyzed several traditional pre-processing methods involving fast Independent Component Analysis (FastICA), Information Maximization (Infomax), and Second-order Blind Source Separation (SOBI). On top of these methods, a SOBI-WT method based on the joint use of the SOBI and Wavelet Transform (WT) is proposed. It offered an effective solution for suppressing artifact components while retaining residual informative data. To provide a comprehensive comparative analysis, these pre-processing methods were applied to eliminate the intra-artifacts and the processed signals were fed to two ready-to-use deep learning models, namely two-step hierarchical neural network (THNN) and SimpleSleepNet for automatic sleep staging. The evaluation was performed on two widely used public datasets, Montreal Archive of Sleep Studies (MASS) and Sleep-EDF Expanded, and a clinical dataset that was collected in Huashan Hospital of Fudan University, Shanghai, China (HSFU). The proposed SOBI-WT method increased the accuracy from 79.0% to 81.3% on MASS, 83.3% to 85.7% on Sleep-EDF Expanded, and 75.5% to 77.1% on HSFU compared with the raw EEG signal, respectively. Experimental results demonstrate that the intra-artifacts bring out a masking negative impact on the deep learning-based sleep staging systems and the proposed SOBI-WT method has the best performance in diminishing this negative impact compared with other artifact elimination methods.

Accurate and timely flood forecasting, facilitated by remote sensing technology, is crucial to mitigate the damage and loss of life caused by floods. However, despite years of research, accurate flood prediction still faces numerous challenges, including complex spatiotemporal features and varied flood patterns influenced by multiple variables. Moreover, long-term flood forecasting is always tricky due to the constantly changing conditions of the surrounding environment. In this study, we propose a heterogeneous dynamic temporal graph convolutional network (HD-TGCN) for flood forecasting. Specifically, we designed a dynamic temporal graph convolution module (D-TGCM) to generate a dynamic adjacency matrix by incorporating a multihead self-attention mechanism, enabling our model to capture the dynamic spatiotemporal features of flood data by utilizing temporal graph convolution operations on the dynamic matrix. Furthermore, to reflect the impact of multiple meteorological and hydrological features on the heterogeneity of flood data, we propose a novel approach that utilizes multiple parallel D-TGCMs for processing heterogeneous graph data and implements a fusion mechanism to capture varied flood patterns influenced by multiple variables. Experiments conducted on a real dataset in Wuyuan County, Jiangxi Province, demonstrate that the HD-TGCN outperforms the state-of-the-art flood prediction models in mean absolute error, Nash–Sutcliffe efficiency, and root-mean-square error, with improvements of 80.32%, 0.15%, and 73.99%, respectively, providing a more accurate flood forecasting method that will play a critical role in future flood disaster prevention and control.

Sleep staging is the essential step in sleep quality assessment and sleep disorders diagnosis. However, most current automatic sleep staging approaches use recurrent neural networks (RNN), resulting in a relatively large training burden. Moreover, these methods only extract information of the whole epoch or adjacent epochs, ignoring the local signal variations within epoch. To address these issues, a novel deep learning architecture named segmented attention network (SAN) is proposed in this paper. The architecture can be divided into feature extraction (FE) and time sequence encoder (TSE). The FE module consists of multiple multiscale CNN (MMCNN) and residual squeeze and excitation block (SE block). The former extracts features from multiple equal-length EEG segments and the latter reinforced the features. The TSE module based on a multi-head attention mechanism could capture the temporal information in the features extracted by FE module. Noteworthy, in SAN, we replaced the RNN module with a TSE module for temporal learning and made the network faster. The evaluation of the model was performed on two widely used public datasets, Montreal Archive of Sleep Studies (MASS) and Sleep-EDFX, and one clinical dataset from Huashan Hospital of Fudan University, Shanghai, China (HSFU). The proposed model achieved the accuracy of 85.5%, 86.4%, 82.5% on Sleep-EDFX, MASS and HSFU, respectively. The experimental results exhibited favorable performance and consistent improvements of SAN on different datasets in comparison with the state-of-the-art studies. It also proved the necessity of sleep staging by integrating the local characteristics within epochs and adjacent informative features among epochs.

Deep sleep staging networks have reached top performance on large-scale datasets. However, these models perform poorer when training and testing on small sleep cohorts due to data inefficiency. Transferring well-trained models from large-scale datasets (source domain) to small sleep cohorts (target domain) is a promising solution but still remains challenging due to the domain-shift issue. In this work, an unsupervised domain adaptation approach, domain statistics alignment (DSA), is developed to bridge the gap between the data distribution of source and target domains. DSA adapts the source models on the target domain by modulating the domain-specific statistics of deep features stored in the Batch Normalization (BN) layers. Furthermore, we have extended DSA by introducing cross-domain statistics in each BN layer to perform DSA adaptively (AdaDSA). The proposed methods merely need the well-trained source model without access to the source data, which may be proprietary and inaccessible. DSA and AdaDSA are universally applicable to various deep sleep staging networks that have BN layers. We have validated the proposed methods by extensive experiments on two state-of-the-art deep sleep staging networks, DeepSleepNet+ and U-time. The performance was evaluated by conducting various transfer tasks on six sleep databases, including two large-scale databases, MASS and SHHS, as the source domain, four small sleep databases as the target domain. Thereinto, clinical sleep records acquired in Huashan Hospital, Shanghai, were used. The results show that both DSA and AdaDSA could significantly improve the performance of source models on target domains, providing novel insights into the domain generalization problem in sleep staging tasks.

In this paper, a novel two-dimensional (2D) generalized optical spatial modulation (GOSM) scheme is proposed for multiple-input multiple-output optical wireless communication (MIMO-OWC) systems. By grouping multiple successive time slots as one time block, 2D GOSM mapping can be performed not only in the space domain but also in the time domain. Specifically, two types of 2D GOSM mapping schemes are designed, including 2D-1 and 2D-2 GOSM mappings. Moreover, to address the high complexity issue of optimal joint maximum-likelihood (ML) detection and the noise amplification and error propagation issues of zero-forcing-based ML (ZF-ML) detection, a deep neural network (DNN)-aided detection scheme is further designed for 2D GOSM systems. Simulation results demonstrate the superiority of the proposed 2D GOSM scheme with deep learning-aided detection for high-speed and low-complexity MIMO-OWC systems. More specifically, a remarkable 3.4-dB signal-to-noise ratio (SNR) gain can be achieved by 2D GOSM in comparison to the conventional one-dimensional (1D) GOSM, when applying the DNN-aided detection.

Sleep staging is the essential step in sleep quality assessment and sleep disorders diagnosis. However, most current automatic sleep staging approaches use recurrent neural networks (RNN), resulting in a relatively large training burden. Moreover, these methods only extract information of the whole epoch or adjacent epochs, ignoring the local signal variations within epoch. To address these issues, a novel deep learning architecture named segmented attention network (SAN) is proposed in this paper. The architecture can be divided into feature extraction (FE) and time sequence encoder (TSE). The FE module consists of multiple multiscale CNN (MMCNN) and residual squeeze and excitation block (SE block). The former extracts features from multiple equal-length EEG segments and the latter reinforced the features. The TSE module based on a multi-head attention mechanism could capture the temporal information in the features extracted by FE module. Noteworthy, in SAN, we replaced the RNN module with a TSE module for temporal learning and made the network faster. The evaluation of the model was performed on two widely used public datasets, Montreal Archive of Sleep Studies (MASS) and Sleep-EDFX, and one clinical dataset from Huashan Hospital of Fudan University, Shanghai, China (HSFU). The proposed model achieved the accuracy of 85.5%, 86.4%, 82.5% on Sleep-EDFX, MASS and HSFU, respectively. The experimental results exhibited favorable performance and consistent improvements of SAN on different datasets in comparison with the state-of-the-art studies. It also proved the necessity of sleep staging by integrating the local characteristics within epochs and adjacent informative features among epochs. [ABSTRACT FROM AUTHOR]

Magnetic Resonance Imaging (MRI) has been proven to be an efficient way to diagnose Alzheimer’s disease (AD). Recent dramatic progress on deep learning greatly promotes the MRI analysis based on data-driven CNN methods using a large-scale longitudinal MRI dataset. However, most of the existing MRI datasets are fragmented due to unexpected quits of volunteers. To tackle this problem, we propose a novel Temporal Recurrent Generative Adversarial Network (TR-GAN) to complete missing sessions of MRI datasets. Unlike existing GAN-based methods, which either fail to generate future sessions or only generate fixed-length sessions, TR-GAN takes all past sessions to recurrently and smoothly generate future ones with variant length. Specifically, TR-GAN adopts recurrent connection to deal with variant input sequence length and flexibly generate future variant sessions. Besides, we also design a multiple scale & location (MSL) module and a SWAP module to encourage the model to better focus on detailed information, which helps to generate high-quality MRI data. Compared with other popular GAN architectures, TR-GAN achieved the best performance in all evaluation metrics of two datasets. After expanding the Whole MRI dataset, the balanced accuracy of AD vs. cognitively normal (CN) vs. mild cognitive impairment (MCI) and stable MCI vs. progressive MCI classification can be increased by 3.61% and 4.00%, respectively. [ABSTRACT FROM AUTHOR]

Fully-supervised deep learning segmentation models are inflexible when encountering new unseen semantic classes and their fine-tuning often requires significant amounts of annotated data. Few-shot semantic segmentation (FSS) aims to solve this inflexibility by learning to segment an arbitrary unseen semantically meaningful class by referring to only a few labeled examples, without involving fine-tuning. State-of-the-art FSS methods are typically designed for segmenting natural images and rely on abundant annotated data of training classes to learn image representations that generalize well to unseen testing classes. However, such a training mechanism is impractical in annotation-scarce medical imaging scenarios. To address this challenge, in this work, we propose a novel self-supervised FSS framework for medical images, named SSL-ALPNet, in order to bypass the requirement for annotations during training. The proposed method exploits superpixel-based pseudo-labels to provide supervision signals. In addition, we propose a simple yet effective adaptive local prototype pooling module which is plugged into the prototype networks to further boost segmentation accuracy. We demonstrate the general applicability of the proposed approach using three different tasks: organ segmentation of abdominal CT and MRI images respectively, and cardiac segmentation of MRI images. The proposed method yields higher Dice scores than conventional FSS methods which require manual annotations for training in our experiments. [ABSTRACT FROM AUTHOR]

Rain streaks in the air show diverse characteristics with different shapes, directions, densities, even the complex overlapped phenomenon, causing great challenges for the deraining task. Recently, deep learning based image deraining methods have been extensively investigated due to their excellent performance. However, most of the existing algorithms still have limitations in removing rain streaks while preserving rich textural details under complicated rain conditions. To this end, we propose to decompose rain streaks into multiple rain layers and individually estimate each of them along the network stages to cope with the increasing abstracts. To better characterize rain layers, an improved non-local block is designed to exploit the self-similarity of rain information by learning the holistic spatial feature correlations while reducing the calculation complexity. Moreover, a mixed attention mechanism is applied to guide the fusion of rain layers by focusing on the local and global overlaps among these rain layers. Extensive experiments on both synthetic rainy/rain-haze/raindrop datasets, real-world samples, the haze, and low-light scenarios show substantial improvements both on quantitative indicators and visual effects over the current state-of-the-art technologies. The source code is available at https://github.com/kuihua/IADN. [ABSTRACT FROM AUTHOR]

In recent years, deep learning has been widely used in breast cancer diagnosis, and many high-performance models have emerged. However, most of the existing deep learning models are mainly based on static breast ultrasound (US) images. In actual diagnostic process, contrast-enhanced ultrasound (CEUS) is a commonly used technique by radiologists. Compared with static breast US images, CEUS videos can provide more detailed blood supply information of tumors, and therefore can help radiologists make a more accurate diagnosis. In this paper, we propose a novel diagnosis model based on CEUS videos. The backbone of the model is a 3D convolutional neural network. More specifically, we notice that radiologists generally follow two specific patterns when browsing CEUS videos. One pattern is that they focus on specific time slots, and the other is that they pay attention to the differences between the CEUS frames and the corresponding US images. To incorporate these two patterns into our deep learning model, we design a domain-knowledge-guided temporal attention module and a channel attention module. We validate our model on our Breast-CEUS dataset composed of 221 cases. The result shows that our model can achieve a sensitivity of 97.2% and an accuracy of 86.3%. In particular, the incorporation of domain knowledge leads to a 3.5% improvement in sensitivity and a 6.0% improvement in specificity. Finally, we also prove the validity of two domain knowledge modules in the 3D convolutional neural network (C3D) and the 3D ResNet (R3D). [ABSTRACT FROM AUTHOR]

Deep-learning-based methods, such as convolution neural network (CNN), have demonstrated their efficiency in hyperspectral image (HSI) classification. These methods can automatically learn spectral-spatial discriminative features within local patches. However, for each pixel in an HSI, it is not only related to its nearby pixels but also has connections to pixels far away from itself. Therefore, to incorporate the long-range contextual information, a deep fully convolutional network (FCN) with an efficient nonlocal module, named ENL-FCN, is proposed for HSI classification. In the proposed framework, a deep FCN considers an entire HSI as input and extracts spectral-spatial information in a local receptive field. The efficient nonlocal module is embedded in the network as a learning unit to capture the long-range contextual information. Different from the traditional nonlocal neural networks, the long-range contextual information is extracted in a specially designed criss-cross path for computation efficiency. Furthermore, using a recurrent operation, each pixel’s response is aggregated from all pixels of HSI. The benefits of our proposed ENL-FCN are threefold: 1) the long-range contextual information is incorporated effectively; 2) the efficient module can be freely embedded in a deep neural network in a plug-and-play fashion; and 3) it has much fewer learning parameters and requires less computational resources. The experiments conducted on three popular HSI data sets demonstrate that the proposed method achieves state-of-the-art classification performance with lower computational cost in comparison with several leading deep neural networks for HSI. [ABSTRACT FROM AUTHOR]

In this paper, we introduce a deep learning aided constraint encoding method to tackle the frequency-constraint microgrid scheduling problem. The nonlinear function between system operating condition and frequency nadir is approximated by using a neural network, which admits an exact mixed-integer formulation (MIP). This formulation is then integrated with the scheduling problem to encode the frequency constraint. With the stronger representation power of the neural network, the resulting commands can ensure adequate frequency response in a realistic setting in addition to islanding success. The proposed method is validated on a modified 33-node system. Successful islanding with a secure response is simulated under the scheduled commands using a detailed three-phase model in Simulink. The advantages of our model are particularly remarkable when the inertia emulation functions from wind turbine generators are considered. [ABSTRACT FROM AUTHOR]

Rear-end collision prediction has gained an increasing attention for safety improvement in smart cities. It is urgent to design efficient warning strategies for rear-end collisions which is one of the main causes of traffic accidents. The existing researches have been conducted to predict the collisions. Learning-based methods are proposed to solve this complicated issue, which the traditional methods are difficult to solve. However, due to some limitations in terms of the feature extraction and prediction performance, back-propagation learning methods are facing challenge. In this paper, we proposed a novel Rear-end Collision Prediction Mechanism with deep learning method (RCPM), in which a convolutional neural network model is established. In RCPM, the dataset is smoothed and expanded based on genetic theory to alleviate the class imbalance problem. The preprocessed dataset is divided into training and testing sets as the input to train our convolutional neural network model. The experimental results show that compared with the Honda, Berkeley and multi-layer perception neural-network-based algorithms, RCPM effectively improves performance to predict rear-end collisions. [ABSTRACT FROM AUTHOR]

In zero-shot learning, attributes play as a bridge from original images to class labels. Therefore, to achieve accurate zero-shot image classification, we mainly focus on improving attribute prediction accuracy by taking full advantage of prior information about attribute from two aspects. First, we present a new attribute classifier called deep attribute prediction (DeepAP) model by using supervised deep convolutional neural networks (DCNNs), where the attribute label information participates in the training of DCNNs. Unlike common DCNNs that are usually used to extract image features, the constructed DCNNs are used to directly predict attribute values from the original input images. Thus, the designed DeepAP model can serve as the mapping from low-level image features to high-level semantic attributes in the traditional direct attribute prediction (DAP) model. Second, another prior information about attribute, i.e., class-attribute matrix is used to mine the attribute-class correlation with sparse representation coefficients. Since the attribute-class correlation can reflect different contributions of attributes to classification, we use it to define attribute weights and incorporate the idea of weighted attributes into DeepAP to form the deep weighted attribute prediction (DWAP) model. Experiments on three real datasets show that DWAP outperforms the deep attribute network and DAP on attribute prediction and zero-shot image classification. [ABSTRACT FROM AUTHOR]

The problem of noninstrusive load monitoring (NILM) is usually formulated as a single-channel blind source separation task, whose successful solution enable fast and convenient load identification and energy disaggregation. When applied at test time, NILM algorithms aim to identify the operating characteristics of individual appliances from an aggregate power measurement of the entire house. Recent advances in deep learning gave rise to many methods that mostly focus on learning a direct mapping from aggregate measurement to individual appliance power. However, these methods are not only computationally expensive, but they often suffer from overfitting and do not generalize very well. In this article, we propose a novel NILM method that leverages advances in statistical learning that have not been properly applied in this domain before. The proposed method consists of three stages: first, a Bayesian nonparametric learning-based approach for appliance state extraction; second, synthetic minority oversampling technique for data augmentation and mitigating the heavy imbalance in switching events; and third, appliance-specific lightweight long short-term memory networks for status classification for each appliance. We adopt a “differential” input (the difference before and after the switching event) to reduce the complexity of network training and make the proposed method robust to multiappliance switching events. Experiments are conducted to demonstrate the effectiveness of the proposed method, achieving superior performance when compared to recent methods. An ablation study is conducted to demonstrate the effectiveness of each module of our method. Finally, we investigate the quality of generated synthetic samples. [ABSTRACT FROM AUTHOR]

In steerable filters, a filter of arbitrary orientation can be generated by a linear combination of a set of “basis filters.” Steerable properties dominate the design of the traditional filters, e.g., Gabor filters and endow features the capability of handling spatial transformations. However, such properties have not yet been well explored in the deep convolutional neural networks (DCNNs). In this paper, we develop a new deep model, namely, Gabor convolutional networks (GCNs or Gabor CNNs), with Gabor filters incorporated into DCNNs such that the robustness of learned features against the orientation and scale changes can be reinforced. By manipulating the basic element of DCNNs, i.e., the convolution operator, based on Gabor filters, GCNs can be easily implemented and are readily compatible with any popular deep learning architecture. We carry out extensive experiments to demonstrate the promising performance of our GCNs framework, and the results show its superiority in recognizing objects, especially when the scale and rotation changes take place frequently. Moreover, the proposed GCNs have much fewer network parameters to be learned and can effectively reduce the training complexity of the network, leading to a more compact deep learning model while still maintaining a high feature representation capacity. The source code can be found at https://github.com/bczhangbczhang. [ABSTRACT FROM AUTHOR]

The classification of benign and malignant parotid tumors is very crucial for the selection of surgical methods and their prognoses. The wide application of deep learning technology in the field of medical imaging also provides new ideas for the computer-aided diagnosis of parotid gland tumors. In addition, because the pathological types of parotid gland tumors are very complicated and the computed tomography (CT) images of benign and malignant patients are also very similar, some clinicians may misjudge tumors due to a lack of experience, which affects the effect of surgical treatment and prognosis. Therefore, this research proposes using deep learning methods to solve this problem. This study uses the four classic pretraining models of VGG16, InceptionV3, ResNet and DenseNet to classify parotid CT images using transfer learning methods and uses an improved convolutional neural network (CNN) model to classify parotid CT images. The experimental results show that the improved CNN model achieves an accuracy of 97.78%, and its classification performance is better than those of the other four transfer learning methods. It can effectively diagnose benign and malignant parotid tumors and improve the diagnostic accuracy.

With the explosive growth in demand for mobile traffic, one of the promising solutions is to offload cellular traffic to small base stations for better system efficiency. Due to increasing system complexity, network operators are facing severe challenges and looking for machine learning-based solutions. In this work, we propose an energy-aware mobile traffic offloading scheme in the heterogeneous network jointly apply deep Q network (DQN) decision making and advanced traffic demand forecasting. The base station control model is trained and verified on an open dataset from a major telecom operator. The performance evaluation shows that DQN with traffic forecasting outperforms others at all levels of mobile traffic demands. Also, the advantage of accurate traffic prediction is more significant under higher traffic loads.

In this paper, a 3D indoor visible light positioning (VLP) system with fast computation time using received signal strength (RSS) is proposed and experimentally demonstrated. Assisted by the deep learning techniques, the complexity of the trilateration problem is greatly reduced, and the trilateration problem can be formulated as a linear mapping leading to faster position estimation than the conventional estimation. Moreover, a new method of off-line preparation is adopted to minimize the workload of the VLP system deployment for more practical usage. The proposition is implemented on an atto-cellular VLP unit, through which the real-time performance and positioning accuracy are demonstrated and validated in a 3D positioning experiment performed in a space of 1.2 × 1.2 × 2 m3. The experimental results show that a positioning accuracy of 11.93 cm in confidence of 90% is achieved with 50 times faster the computation time compared to the conventional scheme.

Deep learning is a growing trend in medical image analysis. There are limited data of deep learning techniques applied in Chest X-rays. This paper proposed a deep learning algorithm for cardiothoracic ratio (CTR) calculation in chest X-rays. A fully convolutional neural network was employed to segment chest X-ray images and calculate CTR. CTR values derived from the deep learning model were compared with the reference standard using Bland-Altman analysis and linear correlation graphs, and intra-class correlation (ICC) analyses. Diagnostic performance of the model for the detection of heart enlargement was assessed and compared with other deep learning methods and radiologists. CTR values derived from the deep learning method showed excellent agreement with the reference standard, with mean difference 0.0004 ± 0.0133, 95% limits of agreement -0.0256 to 0.0264. Correlation coefficient between deep learning and reference standard was 0.965 (P