Search

Abstract Quantification of the cardiac function is vital for diagnosing and curing the cardiovascular diseases. Left ventricular function measurement is the most commonly used measure to evaluate the function of cardiac in clinical practice, how to improve the accuracy of left ventricular quantitative assessment results has always been the subject of research by medical researchers. Although considerable efforts have been put forward to measure the left ventricle (LV) automatically using deep learning methods, the accurate quantification is yet a challenge work as a result of the changeable anatomy structure of heart in the systolic diastolic cycle. Besides, most methods used direct regression method which lacks of visual based analysis. In this work, a deep learning segmentation and regression task-unified network with transformer and spatial–temporal convolution is proposed to segment and quantify the LV simultaneously. The segmentation module leverages a U-Net like 3D Transformer model to predict the contour of three anatomy structures, while the regression module learns spatial–temporal representations from the original images and the reconstruct feature map from segmentation path to estimate the finally desired quantification metrics. Furthermore, we employ a joint task loss function to train the two module networks. Our framework is evaluated on the MICCAI 2017 Left Ventricle Full Quantification Challenge dataset. The results of experiments demonstrate the effectiveness of our framework, which achieves competitive cardiac quantification metric results and at the same time produces visualized segmentation results that are conducive to later analysis.

Abstract The exponential device growth in industrial Internet of things (IIoT) has a noticeable impact on the volume of data generated. Edge-cloud computing cooperation has been introduced to the IIoT to lessen the computational load on cloud servers and shorten the processing time for data. General programmable logic controllers (PLCs), which have been playing important roles in industrial control systems, start to gain the ability to process a large amount of industrial data and share the workload of cloud servers. This transforms them into edge-PLCs. However, the continuous influx of multiple types of concurrent production data streams against the limited capacity of built-in memory in PLCs brings a huge challenge. Therefore, the ability to reasonably allocate memory resources in edge-PLCs to ensure data utilization and real-time processing has become one of the core means of improving the efficiency of industrial processes. In this paper, to tackle dynamic changes in arrival data rate over time at each edge-PLC, we propose to optimize memory allocation with Q-learning distributedly. The simulation experiments verify that the method can effectively reduce the data loss probability while improving the system performance.

Abstract Segmentation of intracranial aneurysm images acquired using magnetic resonance angiography (MRA) is essential for medical auxiliary treatments, which can effectively prevent subarachnoid hemorrhages. This paper proposes an image segmentation model based on a dense convolutional attention U-Net, which fuses deep and rich semantic information with shallow-detail information for adaptive and accurate segmentation of MRA-acquired aneurysm images with large size differences. The U-Net model serves as a backbone, combining dense block and convolution block attention module (CBAM). The dense block is composed of a batch normalization layer, an randomly rectified linear unit activation function, and a convolutional layer, for mitigation of vanishing gradients, for multiplexing of aneurysm features, and for improving the network training efficiency. The CBAM is composed of a channel attention module and a spatial attention module, improving the segmentation performance of feature discrimination and enhancing the acquisition of key feature information. Owing to the large variation of aneurysm sizes, multi-scale fusion is performed during up-sampling, for adaptive segmentation of MRA-acquired aneurysm images. The model was tested on the MICCAI 2020 ADAM dataset, and its generalizability was validated on the clinical aneurysm dataset (aneurysm sizes: 7 mm) supplied by the Affiliated Hospital of Qingdao University. A good clinical application segmentation performance was demonstrated.

Abstract The accurate and robust automatic segmentation of cardiac structures in magnetic resonance imaging (MRI) is significant in calculating cardiac clinical functional indices, and diagnosing heart diseases. Most U‐Net based methods use pooling, transposed convolution, and skip connection operations to integrate the multiscale features for improved segmentation in cardiac MRI. However, this architecture lacks adequate semantic connection between the channel and spatial information, and robustness in segmenting objects with significant shape variations. In this paper, a new multiscale feature attentive U‐Net for cardiac MRI structural segmentation method is proposed. An attention mechanism is adopted after concatenating the multi‐level features to aggregate different scale features and determine on which features to focus. Cascade and parallel dilated convolution is also employed in the decoder blocks and skip connection is employed to enhance the ability of sensing receptive fields for multiscale context information. Furthermore, deep supervision approach with a loss function that combines the dice and cross‐entropy losses to reduce overfitting and ensure better prediction is introduced. The proposed method was evaluated on three public cardiac datasets. The experimental results indicate that the method achieved competitive segmentation performance with the three datasets, which verifies the robustness and generalisability of the proposed network. In comparison with conventional U‐Net methods, the model leverages attention mechanism and dilated convolution block, which increases the semantic connection between the channel and the spatial information, and improves the robustness of the right ventricle segmentation performance. From the view of the Dice scores and segmentation results, the multiscale feature attentive U‐Net method is one of effective methods in segmenting cardiac MRI structures.

Abstract Pulmonary nodules are the main manifestation of early lung cancer. Therefore, accurate detection of nodules in CT images is vital for lung cancer diagnosis. A 3D automatic detection system of pulmonary nodules based on multi-scale attention networks is proposed in this paper to use multi-scale features of nodules and avoid network over-fitting problems. The system consists of two parts, nodule candidate detection (determining the locations of candidate nodules), false positive reduction (minimizing the number of false positive nodules). Specifically, with Res2Net structure, using pre-activation operation and convolutional quadruplet attention module, the 3D multi-scale attention block is designed. It makes full use of multi-scale information of pulmonary nodules by extracting multi-scale features at a granular level and alleviates over-fitting by pre-activation. The U-Net-like encoder-decoder structure is combined with multi-scale attention blocks as the backbone network of Faster R-CNN for detection of candidate nodules. Then a 3D deep convolutional neural network based on multi-scale attention blocks is designed for false positive reduction. The extensive experiments on LUNA16 and TianChi competition datasets demonstrate that the proposed approach can effectively improve the detection sensitivity and control the number of false positive nodules, which has clinical application value.

Surface flattening plays an important role in the whole process of garment design. We proposed a novel method by using three-dimensional triangle mesh flattening in this study. First, the three-dimensional triangle mesh is flattened to a two-dimensional plane to approximate the original surface. The initial flattening results are then used as preliminary guesses for subsequent optimizations. Considering that the deformation energy in the real woven fabric is related to tensile or shear deformation, a simplified fabric deformation model based on energy is proposed to update the energy distribution to determine the best two-dimensional pattern. An innovative unified axis system process is proposed to obtain the deformation energy, and energy relaxation in local flattening is proposed to release the distortion of flattening. Finally, the experimental results show that complex surfaces such as garments could achieve better flattening results. Compared with other energy-based methods in garment design, our proposed methods are more flexible and practical.

Abstract Background Retinal blood vessel segmentation has an important guiding significance for the analysis and diagnosis of cardiovascular diseases such as hypertension and diabetes. But the traditional manual method of retinal blood vessel segmentation is not only time-consuming and laborious but also cannot guarantee the accuracy and efficiency of diagnosis. Therefore, it is especially significant to create a computer-aided method of automatic and accurate retinal vessel segmentation. Methods In order to extract the blood vessels’ contours of different diameters to realize fine segmentation of retinal vessels, we propose a Bidirectional Symmetric Cascade Network (BSCN) where each layer is supervised by vessel contour labels of specific diameter scale instead of using one general ground truth to train different network layers. In addition, to increase the multi-scale feature representation of retinal blood vessels, we propose the Dense Dilated Convolution Module (DDCM), which extracts retinal vessel features of different diameters by adjusting the dilation rate in the dilated convolution branches and generates two blood vessel contour prediction results by two directions respectively. All dense dilated convolution module outputs are fused to obtain the final vessel segmentation results. Results We experimented the three datasets of DRIVE, STARE, HRF and CHASE_DB1, and the proposed method reaches accuracy of 0.9846/0.9872/0.9856/0.9889 and AUC of 0.9874/0.9941/0.9882/0.9874 on DRIVE, STARE, HRF and CHASE_DB1. Conclusions The experimental results show that compared with the state-of-art methods, the proposed method has strong robustness, it not only avoids the adverse interference of the lesion background but also detects the tiny blood vessels at the intersection accurately.

Unstructured point clouds are a representative shape representation of real-world scenes in 3D vision and graphics. Incompletion inevitably arises, due to the way the set of unorganized points is captured, e.g., as fusion of depth images, merged laser scans, or structure-from-x. In this paper, an end-to-end sparse-to-dense multi-encoder neural network (termed an SDME-Net) is proposed for uniformly completing an unstructured point cloud with its shape details preserved. Unlike most existing learning-based shape completion methods that are enforced on the representations of 2D images and 3D voxelization of point clouds, and require priors of the underlying shape's structures, topologies and annotations, the SDME-Net is implemented on the incomplete and even noisy point cloud without any transformation, and makes no specific assumptions about the incompletion distribution and geometry features in the input. Specifically, the defective point cloud is completed and optimized in a sparse-to-dense manner of two-stages. In the first stage, we generate a sparse but complete point cloud based on a bistratal PointNet, and in the second stage, we yield a dense and high-fidelity point cloud by encoding and decoding the sparse result in the first stage using PointNet++. Meanwhile, we combine the distance loss and repulsion loss to generate more uniformly distributed output point clouds closer to the ground-truth counterparts. Qualitative and quantitative experiments on the public ShapeNet dataset illustrate that our approach outperforms the state-of-art learning-based point cloud shape completion methods in terms of real structure recovery, uniformity, and noise/partiality robustness.

In the domain of geotechnical engineering analysis, fast Lagrangian analysis based on the finite difference method is the most commonly used numerical analysis method. It is a classical numerical algorithm applied to calculate the in situ stress. In this paper, we propose a deep learning (DL) architecture called the enhance-and-split feature capsule network embedded in fully convolutional neural networks (ES-Caps-FCN) to predict the in situ stress for a strain-softening model when using the finite difference method for the numerical computation. Experiments indicate that this novel approach is stable and convergent. Compared with some classical prediction methods including linear regression analysis and a deep neural network, the mean squared error of our proposed algorithm is as low as 0.059866%, which is lower than the 0.616676% of the deep neural network prediction algorithm and the 0.978495% of the conventional machine learning algorithm. Additionally, the calculation efficiency of fully trained deep learning models is superior to that of the conventional finite difference method. Therefore, DL is a feasible and promising fast and accurate surrogate for the finite difference method for solving the in situ stress.

Falls are abnormal activity events that occur infrequently; however, they are serious health problems among elderly individuals. With the advancements of technologies, falls have been widely studied by scientific researchers to minimize serious consequences and negative impacts. Fall detection and fall prevention are two strategies to tackle fall issues with a variety of sensing techniques and classifier models. Currently, many reviews on fall-related technologies have been presented and analyzed; however, most of them give surveys on the subfield of fall-related systems, while others are not extensive and comprehensive reviews. In fact, the latest researches have a new trend of fusion-based methods to improve the performance of the fall-related systems based on a combination of different sensors or classifier models. Adaptive threshold and radio frequency-based systems are also researched and proposed recently, which are seldom mentioned in other reviews. Therefore, a global taxonomy for current fall-related studies from four aspects, including current literature reviews, fall detection, and prevention systems based on different sensor apparatus and analytic algorithm, low power techniques, and sensor placements for fall-related systems are conducted in this paper. Several research challenges and issues in the fall-related field are also discussed and analyzed. The objective of this review paper is to conclude and provide a good position of current fall-related studies to inspire researchers in this field.