Coronaviruses are important human and animal pathogens. To date the novel COVID-19 coronavirus is rapidly spreading worldwide and subsequently threatening health of billions of humans. Clinical studies have shown that most COVID-19 patients suffer from the lung infection. Although chest CT has been shown to be an effective imaging technique for lung-related disease diagnosis, chest Xray is more widely available due to its faster imaging time and considerably lower cost than CT. Deep learning, one of the most successful AI techniques, is an effective means to assist radiologists to analyze the vast amount of chest X-ray images, which can be critical for efficient and reliable COVID-19 screening. In this work, we aim to develop a new deep anomaly detection model for fast, reliable screening. To evaluate the model performance, we have collected 100 chest X-ray images of 70 patients confirmed with COVID-19 from the Github repository. To facilitate deep learning, more data are needed. Thus, we have also collected 1431 additional chest X-ray images confirmed as other pneumonia of 1008 patients from the public ChestX-ray14 dataset. Our initial experimental results show that the model developed here can reliably detect 96.00% COVID-19 cases (sensitivity being 96.00%) and 70.65% non-COVID-19 cases (specificity being 70.65%) when evaluated on 1531 Xray images with two splits of the dataset.
Activation functions play a key role in providing remarkable performance in deep neural networks, and the rectified linear unit (ReLU) is one of the most widely used activation functions. Various new activation functions and improvements on ReLU have been proposed, but each carry performance drawbacks. In this paper, we propose an improved activation function, which we name the natural-logarithm-rectified linear unit (NLReLU). This activation function uses the parametric natural logarithmic transform to improve ReLU and is simply defined as. NLReLU not only retains the sparse activation characteristic of ReLU, but it also alleviates the "dying ReLU" and vanishing gradient problems to some extent. It also reduces the bias shift effect and heteroscedasticity of neuron data distributions among network layers in order to accelerate the learning process. The proposed method was verified across ten convolutional neural networks with different depths for two essential datasets. Experiments illustrate that convolutional neural networks with NLReLU exhibit higher accuracy than those with ReLU, and that NLReLU is comparable to other well-known activation functions. NLReLU provides 0.16% and 2.04% higher classification accuracy on average compared to ReLU when used in shallow convolutional neural networks with the MNIST and CIFAR-10 datasets, respectively. The average accuracy of deep convolutional neural networks with NLReLU is 1.35% higher on average with the CIFAR-10 dataset.
Recently, Attention-Gated Convolutional Neural Networks (AGCNNs) perform well on several essential sentence classification tasks and show robust performance in practical applications. However, AGCNNs are required to set many hyperparameters, and it is not known how sensitive the model's performance changes with them. In this paper, we conduct a sensitivity analysis on the effect of different hyperparameters s of AGCNNs, e.g., the kernel window size and the number of feature maps. Also, we investigate the effect of different combinations of hyperparameters settings on the model's performance to analyze to what extent different parameters settings contribute to AGCNNs' performance. Meanwhile, we draw practical advice from a wide range of empirical results. Through the sensitivity analysis experiment, we improve the hyperparameters settings of AGCNNs. Experiments show that our proposals achieve an average of 0.81% and 0.67% improvements on AGCNN-NLReLU-rand and AGCNN-SELU-rand, respectively; and an average of 0.47% and 0.45% improvements on AGCNN-NLReLU-static and AGCNN-SELU-static, respectively.
Automated skin lesion segmentation on dermoscopy images is an essential and challenging task in the computer-aided diagnosis of skin cancer. Despite their prevalence and relatively good performance, deep learning based segmentation methods require a myriad number of training images with pixel-level dense annotation, which is hard to obtain due to the efforts and costs related to dermoscopy images acquisition and annotation. In this paper, we propose the semi- and weakly supervised directional bootstrapping (SWSDB) model for skin lesion segmentation, which consists of three deep convolutional neural networks: a coarse segmentation network (coarse-SN), a dilated classification network (dilated-CN) and an enhanced segmentation network (enhanced-SN). Both the coarse-SN and enhanced-SN are trained using the images with pixel-level annotation, and the dilated-CN is trained using the images with image-level class labels. The coarse-SN generates rough segmentation masks that provide a prior bootstrapping for the dilated-CN and help it produce accurate lesion localization maps. The maps are then fed into the enhanced-SN to transfer the localization information learned from image-level labels to the enhanced-SN to generate segmentation results. Furthermore, we introduce a hybrid loss that is the weighted sum of a dice loss and a rank loss to the coarse-SN and enhanced-SN, ensuring both networks' good compatibility for the data with imbalanced classes and imbalanced hard-easy pixels. We evaluated the proposed SWSDB model on the ISIC-2017 challenge dataset and PH2 dataset and achieved a Jaccard index of 80.4% and 89.4%, respectively, setting a new record in skin lesion segmentation.
A multi-level deep ensemble (MLDE) model that can be trained in an 'end to end' manner is proposed for skin lesion classification in dermoscopy images. In this model, four pre-trained ResNet-50 networks are used to characterize the multiscale information of skin lesions and are combined by using an adaptive weighting scheme that can be learned during the error back propagation. The proposed MLDE model achieved an average AUC value of 86.5% on the ISIC-skin 2018 official validation dataset, which is substantially higher than the average AUC values achieved by each of four ResNet-50 networks.
Network embedding (NE) is playing a principal role in network mining, due to its ability to map nodes into efficient low-dimensional embedding vectors. However, two major limitations exist in state-of-the-art NE methods: structure preservation and uncertainty modeling. Almost all previous methods represent a node into a point in space and focus on the local structural information, i.e., neighborhood information. However, neighborhood information does not capture the global structural information and point vector representation fails in modeling the uncertainty of node representations. In this paper, we propose a new NE framework, struc2gauss, which learns node representations in the space of Gaussian distributions and performs network embedding based on global structural information. struc2gauss first employs a given node similarity metric to measure the global structural information, then generates structural context for nodes and finally learns node representations via Gaussian embedding. Different structural similarity measures of networks and energy functions of Gaussian embedding are investigated. Experiments conducted on both synthetic and real-world data sets demonstrate that struc2gauss effectively captures the global structural information while state-of-the-art network embedding methods fails to, outperforms other methods on the structure-based clustering task and provides more information on uncertainties of node representations.
The Classification of medical images and illustrations in the literature aims to label a medical image according to the modality it was produced or label an illustration according to its production attributes. It is an essential and challenging research hotspot in the area of automated literature review, retrieval and mining. The significant intra-class variation and inter-class similarity caused by the diverse imaging modalities and various illustration types brings a great deal of difficulties to the problem. In this paper, we propose a synergic deep learning (SDL) model to address this issue. Specifically, a dual deep convolutional neural network with a synergic signal system is designed to mutually learn image representation. The synergic signal is used to verify whether the input image pair belongs to the same category and to give the corrective feedback if a synergic error exists. Our SDL model can be trained 'end to end'. In the test phase, the class label of an input can be predicted by averaging the likelihood probabilities obtained by two convolutional neural network components. Experimental results on the ImageCLEF2016 Subfigure Classification Challenge suggest that our proposed SDL model achieves the state-of-the art performance in this medical image classification problem and its accuracy is higher than that of the first place solution on the Challenge leader board so far.