CXR-RefineDet: Single-Shot Refinement Neural Network for Chest X-Ray Radiograph Based on Multiple Lesions Detection

The workload of radiologists has dramatically increased in the context of the COVID-19 pandemic, causing misdiagnosis and missed diagnosis of diseases. The use of artificial intelligence technology can assist doctors in locating and identifying lesions in medical images. In order to improve the accuracy of disease diagnosis in medical imaging, we propose a lung disease detection neural network that is superior to the current mainstream object detection model in this paper. By combining the advantages of RepVGG block and Resblock in information fusion and information extraction, we design a backbone RRNet with few parameters and strong feature extraction capabilities. After that, we propose a structure called Information Reuse, which can solve the problem of low utilization of the original network output features by connecting the normalized features back to the network. Combining the network of RRNet and the improved RefineDet, we propose the overall network which was called CXR-RefineDet. Through a large number of experiments on the largest public lung chest radiograph detection dataset VinDr-CXR, it is found that the detection accuracy and inference speed of CXR-RefineDet have reached 0.1686 mAP and 6.8 fps, respectively, which is better than the two-stage object detection algorithm using a strong backbone like ResNet-50 and ResNet-101. In addition, the fast reasoning speed of CXR-RefineDet also provides the possibility for the actual implementation of the computer-aided diagnosis system.

Far-field super-resolution ghost imaging with a deep neural network constraint

Ghost imaging (GI) facilitates image acquisition under low-light conditions by single-pixel measurements and thus has great potential in applications in various fields ranging from biomedical imaging to remote sensing. However, GI usually requires a large amount of single-pixel samplings in order to reconstruct a high-resolution image, imposing a practical limit for its applications. Here we propose a far-field super-resolution GI technique that incorporates the physical model for GI image formation into a deep neural network. The resulting hybrid neural network does not need to pre-train on any dataset, and allows the reconstruction of a far-field image with the resolution beyond the diffraction limit. Furthermore, the physical model imposes a constraint to the network output, making it effectively interpretable. We experimentally demonstrate the proposed GI technique by imaging a flying drone, and show that it outperforms some other widespread GI techniques in terms of both spatial resolution and sampling ratio. We believe that this study provides a new framework for GI, and paves a way for its practical applications.

Optimization of Historic Building Survey Technology under Artificial Intelligence Wireless Network Technology Environment

In order to optimize the technology of the building, the damage identification of the building structure is studied. Firstly, back propagation neural network (BPNN) and information fusion technology are used to build neural network models. Secondly, the established model is trained. Finally, the displacement mode, natural frequency, Modal Assurance Criterion (MAC), and three kinds of information fusion with only one characteristic information are used as input data to analyse the results of BPNN identification damage. The results show that when the natural frequency is used as the sensitive feature of damage, the accuracy is the highest. The difference between the network output value and the expected value is the smallest, the network output is the most stable, and the network recognition effect is the best. The network output of a mixture of two damage depths is compared with the output of a single damage depth. The data of the network training set composed of the feature data with damage depth of 20 mm and 5 mm has higher accuracy and more accurate damage recognition. This research provides a reference for the optimization of building survey technology and has certain practical value.

A new neural network approach to density calculation on ceramic materials

The materials’ consolidation, especially ceramics, is very important in advanced research development and industrial technologies. Science of sintering with all incoming novelties is the base of all these processes. A very important question in all of this is how to get the more precise structure parameters within the morphology of different ceramic materials. In that sense, the advanced procedure in collecting precise data in submicro-processes is also in direction of advanced miniaturization. Our research, based on different electrophysical parameters, like relative capacitance, breakdown voltage, and [Formula: see text], has been used in neural networks and graph theory successful applications. We extended furthermore our neural network back propagation (BP) on sintering parameters’ data. Prognosed mapping we can succeed if we use the coefficients, implemented by the training procedure. In this paper, we continue to apply the novelty from the previous research, where the error is calculated as a difference between the designed and actual network output. So, the weight coefficients contribute in error generation. We used the experimental data of sintered materials’ density, measured and calculated in the bulk, and developed possibility to calculate the materials’ density inside of consolidated structures. The BP procedure here is like a tool to come down between the layers, with much more precise materials’ density, in the points on morphology, which are interesting for different microstructure developments and applications. We practically replaced the errors’ network by density values, from ceramic consolidation. Our neural networks’ application novelty is successfully applied within the experimental ceramic material density [Formula: see text] [kg/m3], confirming the direction way to implement this procedure in other density cases. There are many different mathematical tools or tools from the field of artificial intelligence that can be used in such or similar applications. We choose to use artificial neural networks because of their simplicity and their self-improvement process, through BP error control. All of this contributes to the great improvement in the whole research and science of sintering technology, which is important for collecting more efficient and faster results.

A multi-modal and multi-scale emotion-enhanced inference model based on fuzzy recognition

Only the label corresponding to the maximum value of the fully connected layer is used as the output category when a neural network performs classification tasks. When the maximum value of the fully connected layer is close to the sub-maximum value, the classification obtained by considering only the maximum value and ignoring the sub-maximum value is not completely accurate. To reduce the noise and improve classification accuracy, combining the principles of fuzzy reasoning, this paper integrates all the output results of the fully connected layer with the emotional tendency of the text based on the dictionary to establish a multi-modal fuzzy recognition emotion enhancement model. The provided model considers the enhancement effect of negative words, degree adverbs, exclamation marks, and question marks based on the smallest subtree on the emotion of emotional words, and defines the global emotional membership function of emojis based on the corpus. Through comparing the results of CNN, LSTM, BiLSTM and GRU on Weibo and Douyin, it is shown that the provided model can effectively improve the text emotion recognition when the neural network output result is not clear, especially for long texts.

Learning shapes cortical dynamics to enhance integration of relevant sensory input

Adaptive sensory behavior is thought to depend on processing in recurrent cortical circuits, but how dynamics in these circuits shapes the integration and transmission of sensory information is not well understood. Here, we study neural coding in recurrently connected networks of neurons driven by sensory input. We show analytically how information available in the network output varies with the alignment between feedforward input and the integrating modes of the circuit dynamics. In light of this theory, we analyzed neural population activity in the visual cortex of mice that learned to discriminate visual features. We found that over learning, slow patterns of network dynamics realigned to better integrate input relevant to the discrimination task. This realignment of network dynamics could be explained by changes in excitatory-inhibitory connectivity amongst neurons tuned to relevant features. These results suggest that learning tunes the temporal dynamics of cortical circuits to optimally integrate relevant sensory input.

A Sustainable Energy Distribution Configuration for Microgrids Integrated to the National Grid Using Back-to-Back Converters in a Renewable Power System

A desire to produce power in microgrids has grown as the demand for electricity has expanded and the cost of installing modern transmission lines over long distances has become infeasible. As such, microgrids pose DC/AC harmonic distortion losses to the voltage supply that eventually fluctuate the output voltage. The key takeaways that this study presents are: (a) a configuration for microgrids integrated to the national grid using back-to-back converters in a renewable power system is achieved; (b) different scenarios of various schemes of sustainability of the power management in microgrids are analyzed; and (c) the reliable and stable network output power distribution is achieved. In this, the proposed control configuration provides space for construction and stability of the power system with sustainability of the power management. The results show that this current configuration works and stabilizes the network in the shortest time possible, and that the DC connection voltage is regulated and maintains reliable network output despite declining slope controllers, DC power and voltage, and power electronic back-to-back converters. Overall, the simulation results show that the proposed system shows acceptable performance under different scenarios. The accuracy of the results is validated with mathematical formulation simulation using MATLAB software. This system can be utilized in distant regions where there is no power grid or in areas where, despite having a power infrastructure, renewable energies are used to supply the output load for the majority of the day and night.

Noise2Noise Improved by Trainable Wavelet Coefficients for PET Denoising

The significant statistical noise and limited spatial resolution of positron emission tomography (PET) data in sinogram space results in the degradation of the quality and accuracy of reconstructed images. Although high-dose radiotracers and long acquisition times improve the PET image quality, the patients’ radiation exposure increases and the patient is more likely to move during the PET scan. Recently, various data-driven techniques based on supervised deep neural network learning have made remarkable progress in reducing noise in images. However, these conventional techniques require clean target images that are of limited availability for PET denoising. Therefore, in this study, we utilized the Noise2Noise framework, which requires only noisy image pairs for network training, to reduce the noise in the PET images. A trainable wavelet transform was proposed to improve the performance of the network. The proposed network was fed wavelet-decomposed images consisting of low- and high-pass components. The inverse wavelet transforms of the network output produced denoised images. The proposed Noise2Noise filter with wavelet transforms outperforms the original Noise2Noise method in the suppression of artefacts and preservation of abnormal uptakes. The quantitative analysis of the simulated PET uptake confirms the improved performance of the proposed method compared with the original Noise2Noise technique. In the clinical data, 10 s images filtered with Noise2Noise are virtually equivalent to 300 s images filtered with a 6 mm Gaussian filter. The incorporation of wavelet transforms in Noise2Noise network training results in the improvement of the image contrast. In conclusion, the performance of Noise2Noise filtering for PET images was improved by incorporating the trainable wavelet transform in the self-supervised deep learning framework.

Novel sensitivity method for evaluating the first derivative of the feed-forward neural network outputs

Evaluating the exact first derivative of a feedforward neural network (FFNN) output with respect to the input feature is pivotal for performing the sensitivity analysis of the trained neural network with respect to the inputs. In this paper, a novel method is presented that computes the analytical quality first derivative of a trained feedforward neural network output with respect to the input features without the need for backpropagation. To this end, the complex step derivative approximation is illustrated, and its implementation in the framework of the feedforward neural network is described. Artificial datasets are generated, and the efficacy of the proposed method for both regression and classification tasks is demonstrated. The results obtained for the regression task indicated that the proposed method is capable of obtaining analytical quality derivatives, and in the case of the classification task, the least relevant features could be identified.