Recently, numerous deep learning models have been proposed for breast cancer diagnosis using multimodal multi-view ultrasound images. However, their performance could be highly affected by overlooking interactions between different modalities and views. Moreover, existing methods struggle to handle cases where certain modalities or views are missing, which limits their clinical applications. To address these issues, we propose a novel Alignment and Imputation Network (AINet) by integrating 1) alignment and imputation pre-training, and 2) hierarchical fusion fine-tuning. Specifically, in the pre-training stage, cross-modal contrastive learning is employed to align features across different modalities, for effectively capturing inter-modal interactions. To simulate missing modality (view) scenarios, we randomly mask out features and then impute them by leveraging inter-modal and inter-view relationships. Following the clinical diagnosis procedure, the subsequent fine-tuning stage further incorporates modality-level and view-level fusion in a hierarchical manner. The proposed AINet is developed and evaluated on three datasets, comprising 15,223 subjects in total. Experimental results demonstrate that AINet significantly outperforms state-of-the-art methods, particularly in handling missing modalities (views). This highlights its robustness and potential for real-world clinical applications.

Understanding and modeling spatio-temporal dynamic systems are critical for numerous real-world applications, yet existing data-driven methods often suffer from inefficiency due to the complexity of their architectures. To address this challenge, we introduce a novel dual-branch framework, GEST, which synergistically integrates graph-based and spectral methods for scalable and effective dynamic system modeling. GEST employs message passing strategy and a tailored attention-enhanced Graph ODE module for continuous-time temporal evolution. Additionally, a Global Fourier Module is proposed to capture long-range dependencies and global dynamics in the frequency domain. Extensive experiments across diverse dynamic scenarios demonstrate the superior predictive performance of GEST and its adaptability to various dynamic scenes. Our codes are available at https://github.com/Eren-01/ICME_GEST.

Continual reinforcement learning (CRL) empowers RL agents with the ability to learn a sequence of tasks, accumulating knowledge learned in the past and using the knowledge for problemsolving or future task learning. However, existing methods often focus on transferring fine-grained knowledge across similar tasks, which neglects the multi-granularity structure of human cognitive control, resulting in insufficient knowledge transfer across diverse tasks. To enhance coarse-grained knowledge transfer, we propose a novel framework called MT-Core (as shorthand for Multi-granularity knowledge Transfer for Continual reinforcement learning). MT-Core has a key characteristic of multi-granularity policy learning: 1) a coarsegrained policy formulation for utilizing the powerful reasoning ability of the large language model (LLM) to set goals, and 2) a fine-grained policy learning through RL which is oriented by the goals. We also construct a new policy library (knowledge base) to store policies that can be retrieved for multi-granularity knowledge transfer. Experimental results demonstrate the superiority of the proposed MT-Core in handling diverse CRL tasks versus popular baselines.

This paper presents a resource-efficient monocular vision framework for leader-follower formation control in GPS-denied environments. To address the challenges of markerless navigation and dynamic interference, our approach integrates geometry-constrained perception with a dual-loop PID control architecture. The main contributions are: (1) A dynamic inverse projection method that reduces scale drift by 62% through optical flow-verified bounding box normalization; (2) A cascaded PID architecture that decouples distance and angle control, achieving 15 FPS on embedded hardware; (3) An implicit communication paradigm enabling 92% occlusion recovery without explicit data exchange. Experimental results demonstrate a 5.2% mean distance error in the 10-25cm range, sub-centimeter adjustments under varying illumination, and robustness in textured environments. Comparative analysis shows a 31% reduction in tracking error compared to marker-based baselines. These findings highlight the potential of our framework for robust, scalable, and infrastructure-free multi-robot formation control.

The massive scale and diverse types of heterogeneous data lead to increased resource consumption and difficulty in feature extraction in data processing. However, large convolutions may not be able to adapt well to the feature extraction requirements of different types of data. Small convolution kernels can enhance feature extraction capabilities and better handle various features in heterogeneous data, thereby improving the classification performance of the entire algorithm for massive heterogeneous data. Therefore, this study proposes a multi label classification algorithm for massive heterogeneous data based on deep convolutional neural networks. The algorithm constructs an improved deep convolution neural network framework, uses convolution layers to extract features of heterogeneous data, and uses the idea of resolving large convolution integrals into small convolutions to reduce the risk of over fitting. In the pooling layer, a hybrid pooling method of adaptive threshold is used to reduce the dimension of heterogeneous data features extracted from the convolution layer. The dimension reduction results are taken as the input of the full connection layer, and the multi label heterogeneous data is classified by softmax classifier. In addition, the central loss function is used to constrain the loss function of softmax to enhance the multi label classification capability of the network. The experimental results show that when the size of convolution kernel is 5*5 and the number is 9, the proposed method achieves the best performance and the lowest classification loss rate.

Multilabel causal feature selection has attracted extensive attention in recent years. Current multilabel causal feature selection algorithms typically employ existing Markov Blanket (MB) search methods for the initial construction of the MB, followed by further optimization. These methods generally treat labels and features as equally weighted nodes during the MB construction process. However, the search for spouse sets often involves extensive conditional independence (CI) tests, which are time-consuming. Furthermore, they fail to consider the distinct contributions of labels and features to the target nodes. Information theory is often used to evaluate the contributions of nodes. Inspired by this, we carry out a theoretical investigation into the causal relationships within multilabel datasets and propose the mutual information-based multilabel causal feature selection (MI-MCF) method. First, MI-MCF employs MI and conditional MI (CMI) instead of CI test when constructing the MB of labels without incurring significant time overhead. Then, MI-MCF uses MI to compare the contributions of features and labels to the target nodes. This helps identify which nodes should be retained when recovering features hindered by strong label correlation. Finally, MI-MCF eliminates spurious nodes through a symmetry check. Experiments on real-world datasets demonstrate that MI-MCF can autonomously determine the optimal number of selected features and consistently outperform compared methods. The code is available at https://github.com/malinjlu/MI-MCF.

Numerous patch-based methods have recently been proposed for histological image based breast cancer classification. However, their performance could be highly affected by ignoring spatial contextual information in the whole slide image (WSI). To address this issue, we propose a novel hierarchical Graph V-Net by integrating 1) patch-level pre-training and 2) context-based fine-tuning, with a hierarchical graph network. Specifically, a semi-supervised framework based on knowledge distillation is first developed to pre-train a patch encoder for extracting disease-relevant features. Then, a hierarchical Graph V-Net is designed to construct a hierarchical graph representation from neighboring/similar individual patches for coarse-to-fine classification, where each graph node (corresponding to one patch) is attached with extracted disease-relevant features and its target label during training is the average label of all pixels in the corresponding patch. To evaluate the performance of our proposed hierarchical Graph V-Net, we collect a large WSI dataset of 560 WSIs, with 30 labeled WSIs from the BACH dataset (through our further refinement), 30 labeled WSIs and 500 unlabeled WSIs from Yunnan Cancer Hospital. Those 500 unlabeled WSIs are employed for patch-level pre-training to improve feature representation, while 60 labeled WSIs are used to train and test our proposed hierarchical Graph V-Net. Both comparative assessment and ablation studies demonstrate the superiority of our proposed hierarchical Graph V-Net over state-of-the-art methods in classifying breast cancer from WSIs. The source code and our annotations for the BACH dataset have been released at https://github.com/lyhkevin/Graph-V-Net.

Cross-modality magnetic resonance (MR) image synthesis can be used to generate missing modalities from given ones. Existing (supervised learning) methods often require a large number of paired multi-modal data to train an effective synthesis model. However, it is often challenging to obtain sufficient paired data for supervised training. In reality, we often have a small number of paired data while a large number of unpaired data. To take advantage of both paired and unpaired data, in this paper, we propose a Multi-scale Transformer Network (MT-Net) with edge-aware pre-training for cross-modality MR image synthesis. Specifically, an Edge-preserving Masked AutoEncoder (Edge-MAE) is first pre-trained in a self-supervised manner to simultaneously perform 1) image imputation for randomly masked patches in each image and 2) whole edge map estimation, which effectively learns both contextual and structural information. Besides, a novel patch-wise loss is proposed to enhance the performance of Edge-MAE by treating different masked patches differently according to the difficulties of their respective imputations. Based on this proposed pre-training, in the subsequent fine-tuning stage, a Dual-scale Selective Fusion (DSF) module is designed (in our MT-Net) to synthesize missing-modality images by integrating multi-scale features extracted from the encoder of the pre-trained Edge-MAE. Furthermore, this pre-trained encoder is also employed to extract high-level features from the synthesized image and corresponding ground-truth image, which are required to be similar (consistent) in the training. Experimental results show that our MT-Net achieves comparable performance to the competing methods even using 70% of all available paired data. Our code will be released at https://github.com/lyhkevin/MT-Net.

High-dimensional multilabel data have increasingly emerged in many application areas, suffering from two noteworthy issues: instances with high-dimensional features and large-scale labels. Multilabel feature selection methods are widely studied to address the issues. Previous multilabel feature selection methods focus on exploring label correlations to guide the feature selection process, ignoring the impact of latent feature structure on label correlations. In addition, one encouraging property regarding correlations between features and labels is that similar features intend to share similar labels. To this end, a latent structure shared (LSS) term is designed, which shares and preserves both latent feature structure and latent label structure. Furthermore, we employ the graph regularization technique to guarantee the consistency between original feature space and latent feature structure space. Finally, we derive the shared latent feature and label structure feature selection (SSFS) method based on the constrained LSS term, and then, an effective optimization scheme with provable convergence is proposed to solve the SSFS method. Better experimental results on benchmark datasets are achieved in terms of multiple evaluation criteria.