Survival prediction for head and neck cancer (HNC) is crucial for clinical treatment. However, current methods extract features only from the tumors and fail to fully use multimodal data. To address these limitations, we propose a multi-task assisted multimodal and multi-granularity survival prediction network. First, a multi-task assisted learning module performs tumor segmentation while capturing global contextual information and prognostic features from both 2D slices and 3D volumes. Second, a multi-graph feature aggregation module constructs distinct graph representations to fuse contextual information from imaging data and adaptively learn relationships within clinical records. Finally, a multimodal and multi-granularity information interaction module integrates 2D and 3D imaging features with clinical and radiomic data for final survival prediction. Experimental results on three HNC datasets demonstrate that the proposed method achieves superior prognostic performance.

Unsupervised domain adaptation (UDA), as a robust transfer learning strategy that utilizes source domain richly labeled data to solve the target domain unlabeled semantic segmentation, has great potential to tackle the problem of data scarcity and domain shift. However, in the medical domain, most existing methods are limited to image-level constrained alignment by image-to-image translation, and few can simultaneously address the dual domain shift of both modality and lesion discrepancies. The motivation of this paper is to explore the potential of UDA for addressing cross-modal and cross-lesion brain lesion segmentation without generative training. The specific novelty of our proposal mainly includes: (1) proposing a UDA technology for cross-modality and cross-disease brain imaging segmentation, which are more general and urgent in clinical practice; (2) exploring the potential of Fourier-based bi-directional frequency domain adaptation to alleviate imaging protocol discrepancies as an alternative to burdensome generative training; (3) proposing a self-training strategy with category-level prototype contrastive learning to exploit cross-domain similar semantic features. The proposed CMCD-Net exhibited encouraging adaptation performance in two segmentation tasks: multi-modality MRI gliomas, multi-modality MRI glioma to meningioma. In addition, massive ablation experiments validated the effectiveness of each component in our proposal. The code is publicly available at https://github.com/Snow1949/C2_MUDA.

Skin lesion segmentation is a complex and severe project, which aims to accurately segment abnormal regions in skin lesion images. However, obtaining accurate segmentation results is difficult because of the great uncertainty in the shape, location, and scale of the target region. To address these challenges, we propose a higher-order spatial interaction framework with dual cross global efficient attention (DGEAHorNet), which employs a neural network architecture based on recursive gate convolution to adequately extract multi-scale contextual information from images. Specifically, a Dual Cross-Attentions (DCA) is added to the skip connection that can effectively blend multi-stage encoder features and narrow the semantic gap. In the bottleneck stage, global channel spatial attention module (GCSAM) is used to extract image global information. To obtain better feature representation, we feed the output from the GCSAM into the multi-branch dense layer (SENetV2) for excitation. Furthermore, we adopt Depthwise Over-parameterized Convolutional Layer (DO-Conv) in order to replace the common convolutional layer in the input and output part of our network, then add Efficient Attention (EA) to diminish computational complexity and enhance our model’s performance. For evaluating the effectiveness of our proposed DGEAHorNet, we conduct comprehensive experiments on three publicly-available skin datasets, and achieving 0.9320, 0.9337 and 0.9474 in Dice similarity coefficient on ISIC2018, ISIC2017, and PH2 severally. Our proposed method outperforms other state-of-the-art methods. The code is available at https://github.com/penghaixin/mymodel.

Melanoma is a particularly aggressive cancer; however, early detection and treatments can increase patient survival rates. Deep learning techniques have practical applications in segmentation to assist physicians in making early melanoma diagnoses. For the problems of lesion regions with different shapes and sizes, discontinuous and blurred boundaries, high similarity between lesion regions and background in dermoscopic images, a multi-scale feature fusion network (MFF-Net) is proposed in this paper. The network contains an edge enhancement module (EEM) for enhancing the edge features of lesion regions. A multi-branch feature fusion module (MBF) for enhancing the feature extraction capability of the network. More local continuity of the feature map is obtained by a hybrid encoder. Using a global feature fusion module (GFF) in the decoder to learn more contextual information to alleviate the hair interference in skin lesion image segmentation and the problem of low contrast. To validate the effectiveness of the proposed network, we conducted extensive experiments to evaluate it on three public skin lesion segmentation datasets (ISIC2018, PH2, and HAM10000), Dice reached 93.17% on ISIC2018 data set, and IoU reached 93.88% on PH2data set and the comparison with the latest methods also validates the superiority of our proposed MFF-Net in terms of accuracy. Our code on https://github.com/gih23/MFF-Net.

Accurate segmentation of organs and tumours from medical images is to diagnose and treat diseases more accurately. Many organs, such as the representative pancreas and spleen, have blurred boundaries, small size, large variance, and inter-class imbalance. 2.5D methods based on 2D networks are currently widely used in various areas of medical image segmentation. However, 2D network-based methods are still unable to truly balance the contextual residual difference information. Therefore, we propose a new 2.5D multi-path Transformer fusion network(MTr-Net) that combines Z-axis information from 3D networks to address organs and tumours boundary deformation, while balancing contextual residual information. A fusion of two different methods is proposed to further refine the segmentation results of organs and tumours. Our method is evaluated on four widely accepted public datasets of pancreas and tumour, spleen, and skin lesions, which contain different imaging modalities. Specifically, these include the NIH public dataset, the MSD Pancreas public dataset, the MSD Spleen dataset and the ISIC2018 dataset. The algorithm performs well on all the above datasets. Our source codes will be released at https://github.com/graduation37289/MTr-net, once the manuscript is accepted for publication.

Automatic brain tumor segmentation using multi-parametric magnetic resonance imaging (mpMRI) holds substantial importance for brain diagnosis, monitoring, and therapeutic strategy planning. Given the constraints inherent to manual segmentation, adopting deep learning networks for accomplishing accurate and automated segmentation emerges as an essential advancement. In this article, we propose a modality fusion diffractive network (MFD-Net) composed of diffractive blocks and modality feature extractors for the automatic and accurate segmentation of brain tumors. The diffractive block, designed based on Fraunhofer's single-slit diffraction principle, emphasizes neighboring high-confidence feature points and suppresses low-quality or isolated feature points, enhancing the interrelation of features. Adopting a global passive reception mode overcomes the issue of fixed receptive fields. Through a self-supervised approach, the modality feature extractor effectively utilizes the inherent generalization information of each modality, enabling the main segmentation branch to focus more on multimodal fusion feature information. We apply the diffractive block on nn-UNet in the MICCAI BraTS 2022 challenge, ranked first in the pediatric population data and third in the BraTS continuous evaluation data, proving the superior generalizability of our network. We also train separately on the BraTS 2018, 2019, and 2021 datasets. Experiments demonstrate that the proposed network outperforms state-of-the-art methods.

Accurate segmentation of head and neck organs at risk is crucial in radiotherapy. However, the existing methods suffer from incomplete feature mining, insufficient information utilization, and difficulty in simultaneously improving the performance of small and large organ segmentation. In this paper, a multistage hierarchical learning network is designed to fully extract multidimensional features, combined with anatomical prior information and imaging features, using multistage subnetworks to improve the segmentation performance. First, multilevel subnetworks are constructed for primary segmentation, localization, and fine segmentation by dividing organs into two levels-large and small. Different networks both have their own learning focuses and feature reuse and information sharing among each other, which comprehensively improved the segmentation performance of all organs. Second, an anatomical prior probability map and a boundary contour attention mechanism are developed to address the problem of complex anatomical shapes. Prior information and boundary contour features effectively assist in detecting and segmenting special shapes. Finally, a multidimensional combination attention mechanism is proposed to analyze axial, coronal, and sagittal information, capture spatial and channel features, and maximize the use of structural information and semantic features of 3D medical images. Experimental results on several datasets showed that our method was competitive with state-of-the-art methods and improved the segmentation results for multiscale organs.

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.

With the gradual introduction of deep learning into the field of information hiding, the capacity of information hiding has been greatly improved. Therefore, a solution with a higher capacity and a good visual effect had become the current research goal. A novel high-capacity information hiding scheme based on improved U-Net was proposed in this paper, which combined improved U-Net network and multiscale image analysis to carry out high-capacity information hiding. The proposed improved U-Net structure had a smaller network scale and could be used in both information hiding and information extraction. In the information hiding network, the secret image was decomposed into wavelet components through wavelet transform, and the wavelet components were hidden into image. In the extraction network, the features of the hidden image were extracted into four components, and the extracted secret image was obtained. Both the hiding network and the extraction network of this scheme used the improved U-Net structure, which preserved the details of the carrier image and the secret image to the greatest extent. The simulation experiment had shown that the capacity of this scheme was greatly improved than that of the traditional scheme, and the visual effect was good. And compared with the existing similar solution, the network size has been reduced by nearly 60%, and the processing speed has been increased by 20%. The image effect after hiding the information was improved, and the PSNR between the secret image and the extracted image was improved by 6.3 dB.

Ensemble learning is an effective approach for improving the generalization ability of base classifiers. To generate a set of accurate and diverse base classifiers, different data perturbation schemes have been proposed. For instance, Bagging perturbs the training data via bootstrap sampling. However, when a stable learning algorithm (e.g., KNN, Naive Bayes) is used to train base classifiers, the sole perturbation on the training data may not produce diverse base classifiers. In this paper, by using the attribute reduction technology in rough sets, a multi-modal perturbation-based algorithm (called ‘E _ EARBS’) is proposed for the ensemble of base classifiers. E _ EARBS simultaneously perturbs the feature space, training data and learning parameters, where the relative decision entropy(RDE)-based approximate reducts are used to perturb the feature space, and bootstrap sampling is used to perturb the training data. Experimental results show that E _ EARBS can provide competitive solutions for ensemble learning.

Multi-robot services are widely used to improve the efficiency of industry Internet applications, especially in smart factories. Under the situation that the tasks are becoming more and more intensive, how can smart factories use limited robot resources to complete tasks more efficiently? In order to solve this problem, we transform it into a resource-constrained multiproject scheduling problem, and consider using a combinatorial auction method to get the solution. In order to ensure the security of the system and solve the transaction cost of the robot system, we adopt Blockchain technology and smart contracts to organize the work of the robots. We finally conducted performance analysis of our proposed method, and the results show that smart contracts and combined auction algorithms are safe and effective.

Summary Web service recommendation becomes an increasingly important issue when more and more services are published on the Internet. Many Web service recommendation methods have been proposed in recent years, most of which adopted collaborative filtering (CF) techniques. In general, these approaches have two limitations. Firstly, they rarely leverage user ratings since this kind of explicit feedback is difficult to collect for Web services. Secondly, the new user cold‐start problem is an inherent limitation of CF because the new users have not yet cast sufficient numbers of votes. In this paper, pseudo ratings of services constructed based on plenty of user‐service interactions, also known as a kind of implicit feedback, are provided to represent users' preferences on services. Based on these pseudo ratings, we present a novel Web service recommendation approach, which can alleviate the cold‐start problem by integrating contextual information and an online learning model. Experiments conducted on a real world data set show that, compared with the method without contextual information, our proposed approach that handles the cold start problem by integrating contextual information can achieve better F‐Measure performance (5.08 times increase on average). Moreover, the proposed online recommendation approach can dramatically decrease the time overhead while keeping the similar recommendation performance.

Summary Search engine based Web service discovery model suffers from the semantic sparsity problem due to the fact that Web services are described in short texts, which in turn leads to poor recall. To address this issue, external information that enriches the semantics of the Web service and improves discovery performance has been highly concerned. In light of this, we propose a novel Web service discovery approach that uses the neural topic model, which seamlessly integrates tagging information and word embedding for semantic sparsity problem. More specifically, instead of clustering Web services as done in most existing service discovery approaches, we use word embedding to map the words as continuous embeddings to embody external semantics of the service description. We also leverage the neural topic model in service discovery, which takes continuous word distribution as the input and interprets the Web service description as a hierarchical model. Based on the neural topic model and word embedding, we propose an efficient Web service query and ranking approach. Experiments conducted on a real‐world Web service dataset demonstrate the effectiveness of the proposed approach.

Mesh generation is a key step for many areas. Here, an adaptive mesh generation algorithm for piecewise linear domains based on constrained Delaunay is proposed. We focus on improving the quality of the meshes generated and make them conforming to an isotropic size function which is specified by the user. The preliminary mesh with boundaries conforming is generated first. We try to add points to split the mesh either bad shape or not conforming to the size function. The new point can be added to the domain only if it is not close to the existing point in the domain. Finally, local meshes are reconstructed to make all meshes comply with the constrained Delaunay. Experimental comparisons show that the new algorithm is advantageous in generating adaptive mesh. The presented method can be used in a wide range of areas such as computer vision, finite element method and many other areas.

Abstract Virtual endoscopy is a technique to explore hollow organs and anatomical cavities using 3D medical imaging and computer graphics. In this paper, boundary model and local feature structure are used to realize tissue segmentation, and a new efficient algorithm is presented to solve path planning. As to real‐time processing, a frame in virtual endoscopy is divided into near viewpoint part and far viewpoint part based on volume data characteristics in our method. In the aspect of scene rendering, a ray casting algorithm based on the boundary voxel is proposed. Thus the voyage images can be rendered in real time with high quality in virtual endoscopy system by using these techniques. The experiments show that application results of our algorithm in tissue segmentation, path planning, scene rendering are better than other algorithms. Copyright © 2011 John Wiley & Sons, Ltd.