Electroencephalogram (EEG) is susceptible to various nonneural physiological artifacts. Automatic artifact removal from EEG data remains a key challenge for extracting relevant information from brain activities. To adapt to variable subjects and EEG acquisition environments, this paper presents an automatic online artifact removal method based on a priori artifact information. The combination of discrete wavelet transform and independent component analysis (ICA), wavelet-ICA, was utilized to separate artifact components. The artifact components were then automatically identified using a priori artifact information, which was acquired in advance. Subsequently, signal reconstruction without artifact components was performed to obtain artifact-free signals. The results showed that, using this automatic online artifact removal method, there were statistical significant improvements of the classification accuracies in both two experiments, namely, motor imagery and emotion recognition.

No abstract

No abstract

In image classification of deep learning, adversarial examples where inputs intended to add small magnitude perturbations may mislead deep neural networks (DNNs) to incorrect results, which means DNNs are vulnerable to them. Different attack and defense strategies have been proposed to better research the mechanism of deep learning. However, those research in these networks are only for one aspect, either an attack or a defense, not considering that attacks and defenses should be interdependent and mutually reinforcing, just like the relationship between spears and shields. In this paper, we propose Cycle-Consistent Adversarial GAN (CycleAdvGAN) to generate adversarial examples, which can learn and approximate the distribution of original instances and adversarial examples. For CycleAdvGAN, once the Generator A and D are trained, A G can generate adversarial perturbations efficiently for any instance, so as to make DNNs predict wrong, and D G recovery adversarial examples to clean instances, so as to make DNNs predict correct. We apply CycleAdvGAN under semi-white box and black-box settings on two public datasets MNIST and CIFAR10. Using the extensive experiments, we show that our method has achieved the state-of-the-art adversarial attack method and also efficiently improve the defense ability, which make the integration of adversarial attack and defense come true. In additional, it has improved attack effect only trained on the adversarial dataset generated by any kind of adversarial attack. INTRODUCTIONWith the Deep Neural Networks (DNNs) rapid development, they have achieved great successes in various tasks handling the image recognition[1], text processing[2] and speech recognition[3]. Despite the great success, DNNs have been proved to be vulnerable and susceptible to adversarial example[4], the carefully crafted samples looking similar to natural images but designed to mislead a pretrained model. On the one hand, adversarial example leads to potential security threats by attacking or misleading the practical deep learning applications, for example mistaking a stop sign for a yield sign[5] when auto driving, and a thief for a staff when face recognition[6]. On the other hand, adversarial examples are also valuable and beneficial to not only the deep learning models but also the machine learning model, as they can enhance the robust of models and provide insights into their strengths, weaknesses, and blind-spots[7]. CycleAdvGAN: integration of adversarial attack and defence 2 The strategy to generate adversarial examples is to intentionally add imperceptible perturbations to clean instances, for fooling DNNs to make wrong predictions. In the past years, various attack algorithms have been developed to produce adversarial examples in the white-box manner with the knowledge of the structure and parameters of a given model, including gradient based algorithms such as fast gradient sign method[8] and iterative variants of gradient-based methods[9], optimization-based methods such as box-constrained LBFGS[4] and ...

Decoding human brain activities, especially reconstructing human visual stimuli via functional magnetic resonance imaging (fMRI), has gained increasing attention in recent years. However, the high dimensionality and small quantity of fMRI data impose restrictions on satisfactory reconstruction, especially for the reconstruction method with deep learning requiring huge amounts of labelled samples. When compared with the deep learning method, humans can recognize a new image because our human visual system is naturally capable of extracting features from any object and comparing them. Inspired by this visual mechanism, we introduced the mechanism of comparison into deep learning method to realize better visual reconstruction by making full use of each sample and the relationship of the sample pair by learning to compare. In this way, we proposed a Siamese reconstruction network (SRN) method. By using the SRN, we improved upon the satisfying results on two fMRI recording datasets, providing 72.5% accuracy on the digit dataset and 44.6% accuracy on the character dataset. Essentially, this manner can increase the training data about from n samples to 2n sample pairs, which takes full advantage of the limited quantity of training samples. The SRN learns to converge sample pairs of the same class or disperse sample pairs of different class in feature space.

AimsTo evaluate the diagnostic value of three-dimensional rotational angiography (3D-RA) of intracranial micro-aneurysms (diameter ≤ 3 mm) and provide guidance on the value of endovascular treatment. Materials and methods 43 patients with intracranial micro-aneurysms were analyzed retrospectively, all patients had undergone angiography with both conventional 2D-DSA(Two-Dimensional Digital Subtraction Angiography) and rotational angiography with three-dimensional reconstruction; the frequency of detection of aneurysms, depiction of aneurysm neck, radiation dose, and the dosage of contrast agent were recorded respectively. Results 55 pieces of aneurysms were detected out from the 43 cases with intracranial micro-aneurysms by 3D-RA. But only 39 cases were detected out using 2D-DSA from the 55 samples, there were significant differences with regards to detection rate (P < 0.05). There were significant differences in radiation dose and dosage of contrast agent (P < 0.05) between the two methods of using 3D-RA can improve the detection rate of micro-aneurysms, which bestows obvious advantages on displaying the shape of aneurysms, the aneurysm neck at the best angle, and the relationship with the parent artery, at the same time, the amount of contrast agent and radiation dose are reduced in 3D-RA compared to 2D-DSA. Keywords Three-dimensional rotational angiography, Intracranial micro-aneurysm, Three dimensional reconstruction AimsIn order to improve the medical imaging, some immune computation theories and immune algorithms were reviewed and compared. Materials and methodsThe immune computation theories include the self and nonself theory, danger theory, artificial immune network etc. The immune algorithms include self/nonself detection algorithm, normal model construction algorithm, clonal selection algorithm, negative selection algorithm, danger model algorithm and hybrid immune algorithm etc. We improved the clonal selection algorithm to attain the optimal threshold for better segmentation of the medical images than the traditional approach. Results The X-ray medical image of the tuberculosis was processed with the improved clonal selection algorithm and noise filtering, and the output medical image of our approach is better for diagnosis than that of traditional image processing methods. ConclusionsThe immune algorithm can be improved to establish a better medical imaging, and this kind of medical application system is inspired from the human immune system. AcknowledgementsSupported by the project grants from National Natural Science Foundation of China (Grand No. 61673007, 61271114, 11572084, 11472061 and 61203325) Aims Traditional medical image classification methods focus on feature representation and classifier design. However, they seldom concerns data selection used for model training, which plays key role for model tuning and parameter optimization. This paper proposes a novel medical image classification method according to guided bagging. Materials and methods First, unsupervised learning is implemented...

No abstract

Abstract. Emotional illnesses are a kind of typical brain dysfunction disease that influences daily work and urban life. Previous studies have showed that real-time functional magnetic resonance imaging neurofeedback (rt-fMRI NF) had effect on improving the treatment of some mental diseases. The neurofeedback signal usually is the activation of single or multiple brain areas, which limits the applications of NF-based brain network. However, the causal relationships in the emotion regulatory network remain unclear. The amygdala is a complex brain structure that consists of functionally distinct nuclei, particularly the basolateral amygdala (BLA) and centromedial amygdala (CMA). In this study, a group of healthy participants attempted to learn emotional self-regulation of the feedback based on their brain state. We performed granger causality analysis (GCA) for the significant vibrancy in information interactions between the BLA and CMA, both in the emotion regulatory network and among amygdala nuclei. Compared with the CMA, the BLA receives more information flows within the network regions, particularly from the right thalamus to the right rostral anterior cingulate cortex. Comparison of changes in amygdala nuclei shows that the information flow is suppressed from the right BLA to the left CMA. Hence, the BLA plays a critical role in emotion processing and transfer. This study highlights the possible use for rt-fMRI NF-based emotion regulation to the brain network.