Spatial transcriptomics technologies carry out advanced sequencing analysis of molecular profiles with a spatial context, providing multi-source information essential for elucidating biological regulatory mechanisms. Nonetheless, it poses challenges in the integration of raw spatial coordinates with high-dimensional gene expression profiles in their native feature space. While spatial-aware methods effectively aggregate molecular information from local spatial neighborhoods, they fail to explore the long-range relationships associated with gene expression data. To address this issue, this paper introduces a novel approach termed GraphSTAR that encodes both spatial and gene expression data into undirected graphs, characterizing the local spatial proximity and global transcriptional similarity, respectively. Through a graph aggregation process, GraphSTAR integrates these diverse data sources within a joint graph structure, effectively modeling both local neighborhood relationships and long-range functional associations. Subsequently, a reassembled graph neural network is established by incorporating the graph aggregation into the feed-forward propagation using proximal operators, progressively refining spatial-informed latent representation to decipher spatial expression patterns of genes. Extensive experiments on benchmark datasets demonstrate that GraphSTAR outperforms state-of-the-art methods in both spatial domain identification and cell-type annotation tasks.

Text data augmentation is an effective strategy for overcoming the challenge of limited sample sizes in many natural language processing (NLP) tasks. This challenge is especially prominent in the few-shot learning (FSL) scenario, where the data in the target domain is generally much scarcer and of lowered quality. A natural and widely used strategy to mitigate such challenges is to perform data augmentation to better capture data invariance and increase the sample size. However, current text data augmentation methods either can’t ensure the correct labeling of the generated data (lacking faithfulness), or can’t ensure sufficient diversity in the generated data (lacking compactness), or both. Inspired by the recent success of large language models (LLM), especially the development of ChatGPT, we propose a text data augmentation approach based on ChatGPT (named ”AugGPT”). AugGPT rephrases each sentence in the training samples into multiple conceptually similar but semantically different samples. The augmented samples can then be used in downstream model training. Experiment results on multiple few-shot learning text classification tasks show the superior performance of the proposed AugGPT approach over state-of-the-art text data augmentation methods in terms of testing accuracy and distribution of the augmented samples.

The Segment Anything Model (SAM) has gained popularity as a versatile image segmentation method, thanks to its strong generalization capabilities across various domains. However, when applied to optic disc (OD) and optic cup (OC) segmentation tasks, SAM encounters challenges due to the complex structures, low contrast, and blurred boundaries typical of fundus images, leading to suboptimal performance. To over-come these challenges, we introduce a novel model, FunduSAM, which incorporates several Adapters into SAM to create a deep network specifically designed for OD and OC segmentation. The FunduSAM utilizes Adapter into each transformer block after encoder for parameter fine-tuning (PEFT). It enhances SAM’s feature extraction capabilities by designing a Convolutional Block Attention Module (CBAM), addressing issues related to blurred boundaries and low contrast. Given the unique requirements of OD and OC segmentation, polar transformation is used to convert the original fundus OD images into a format better suited for training and evaluating FunduSAM. A joint loss is used to achieve structure preservation between the OD and OC, while accurate segmentation. Extensive experiments on the REFUGE dataset, comprising 1,200 fundus images, demonstrate the superior performance of FunduSAM compared to five mainstream approaches.

The knowledge graph (KG) is a highly needed basis to support the high-fidelity and high-interpretability modeling of various tasks in healthcare artificial intelligence. In this work, we focus on constructing an oncology knowledge graph that will be used in downstream cancer research and solution development. Modern supervised learning for knowledge graph construction requires a large amount of manually labeled data, which makes the process time-consuming and labor-intensive. Although there exists multiple research on named entity recognition and relation extraction based on distantly supervised learning, constructing a domain-specific knowledge graph from large collections of textual data without manual annotations is still an urgent problem to be solved. In response, we propose an integrated framework for adapting and re-learning knowledge graphs from a general domain (biomedical in our case) to a fine-defined domain (oncology). In this framework, we apply distant-supervision on cross-domain knowledge graph adaptation. Consequently, no manual data annotation is required to train the model. We introduce a novel iterative training strategy to facilitate the discovery of domain-specific named entities and triplets. Experimental results indicate that the proposed framework can perform domain adaptation and construction of knowledge graphs efficiently.

Due to the inherent high-dimensional characteristics of genomic data, traditional single metric/kernel-based clustering methods fail to accurately perform data analysis. To address this issue, we propose a multi-kernel clustering with tensor fusion on the Grassmann manifold (MKCTM). Specifically, multiple kernel functions are employed to map data into different kernel spaces and utilize tensor representations to capture their high-order relationships. By introducing a tensor low-rank constraint, we maximize the correlation among kernels while separating the noise and redundancy information from kernel tensor. Finally, the learned kernel tensor is fused on the Grassmann manifold to obtain the final kernel matrix for enhancing clustering. We integrate tensor learning and tensor fusion steps into a unified optimization model and propose an efficient iterative optimization algorithm to solve it. Our proposed method is evaluated on six high-dimensional gene expression datasets against eight popular baseline methods. The remarkable experimental performance demonstrates the exceptional effectiveness of our approach. Our code is available at https://github.com/foureverfei/MKCTM.git

Whole slide image (WSI) analysis represents the current gold standard for cancer diagnosis. To date many fully supervised learning methods have been proposed for WSI classification and segmentation. However, these methods are substantially limited by accurate pixel-level labels, which are labor-intensive to obtain. To solve this problem, we developed an end-to-end multiple instance learning (MIL)-based network for WSI segmentation using coarse-grained labels only. Our network consists of two main components. First, we introduce a hybrid transformer architecture, which uses a fusion mechanism to fuse the feature maps of the convolutional neural network (CNN) and transformer. Second, a novel regional MIL aggregator is proposed, which is used to identify the key instances and address the problem of data imbalance. Unlike the current MIL methods that treat each instance as being independent, our method gathers the information from neighborhood pixels of each instance and captures the correlation between instances. We evaluated our network on CAMELYON16. The benchmarking experiments and ablation studies show that the performance of our method is competitive with those of fully supervised methods and is also better than those of previous MIL segmentation methods.