Leyi Wei

Zhongyu He, Yuanfan Li, Fei Huang et al.

Long-horizon LLM agents can benefit from reusable skills, yet existing skill-based methods often rely on external skill generators during training or persistent skill retrieval at inference, increasing engineering complexity, context length, and deployment latency. We propose Self-Internalizing Reinforcement learning with Intrinsic skills (SIRI), a three-phase framework that enables agents to discover, validate, and internalize skills without external skill generators or inference-time skill banks. SIRI first warms up the policy with GiGPO to acquire basic interaction ability and collect successful skill-free trajectories. It then performs self-skill mining, where the current policy summarizes compact skills from its own successful plain rollouts and validates them through paired skill-augmented and skill-free rollouts. Finally, SIRI distills only beneficial skill-guided action tokens into the plain policy using trajectory-level utility and action-level advantage. At inference, the agent runs with the original prompt only. On ALFWorld and WebShop with Qwen2.5-7B-Instruct, SIRI improves GiGPO from 0.908 to 0.930 on ALFWorld and from 0.728 to 0.813 on WebShop, outperforming prompt-based, RL-based, and memory-augmented baselines. Further analysis shows that our self-mining strategy can achieve performance comparable to distillation with closed-source large model. Our code is available at https://github.com/kirito618/SIRI.

Tianling Liu, Ran Su, Changming Sun et al.

Ovarian cancer is one of the most serious cancers that threaten women around the world. Epithelial ovarian cancer (EOC), as the most commonly seen subtype of ovarian cancer, has rather high mortality rate and poor prognosis among various gynecological cancers. Survival analysis outcome is able to provide treatment advices to doctors. In recent years, with the development of medical imaging technology, survival prediction approaches based on pathological images have been proposed. In this study, we designed a deep framework named EOCSA which analyzes the prognosis of EOC patients based on pathological whole slide images (WSIs). Specifically, we first randomly extracted patches from WSIs and grouped them into multiple clusters. Next, we developed a survival prediction model, named DeepConvAttentionSurv (DCAS), which was able to extract patch-level features, removed less discriminative clusters and predicted the EOC survival precisely. Particularly, channel attention, spatial attention, and neuron attention mechanisms were used to improve the performance of feature extraction. Then patient-level features were generated from our weight calculation method and the survival time was finally estimated using LASSO-Cox model. The proposed EOCSA is efficient and effective in predicting prognosis of EOC and the DCAS ensures more informative and discriminative features can be extracted. As far as we know, our work is the first to analyze the survival of EOC based on WSIs and deep neural network technologies. The experimental results demonstrate that our proposed framework has achieved state-of-the-art performance of 0.980 C-index. The implementation of the approach can be found at https://github.com/RanSuLab/EOCprognosis.

Yu Wang, Chao Pang, Yuzhe Wang et al.

Leveraging artificial intelligence for automatic retrosynthesis speeds up organic pathway planning in digital laboratories. However, existing deep learning approaches are unexplainable, like "black box" with few insights, notably limiting their applications in real retrosynthesis scenarios. Here, we propose MechRetro, a chemical-mechanism-driven graph learning framework for interpretable retrosynthetic prediction and pathway planning, which learns several retrosynthetic actions to simulate a reverse reaction via elaborate self-adaptive joint learning. By integrating chemical knowledge as prior information, we design a novel Graph Transformer architecture to adaptively learn discriminative and chemically meaningful molecule representations, highlighting the strong capacity in molecule feature representation learning. We demonstrate that MechRetro outperforms the state-of-the-art approaches for retrosynthetic prediction with a large margin on large-scale benchmark datasets. Extending MechRetro to the multi-step retrosynthesis analysis, we identify efficient synthetic routes via an interpretable reasoning mechanism, leading to a better understanding in the realm of knowledgeable synthetic chemists. We also showcase that MechRetro discovers a novel pathway for protokylol, along with energy scores for uncertainty assessment, broadening the applicability for practical scenarios. Overall, we expect MechRetro to provide meaningful insights for high-throughput automated organic synthesis in drug discovery.

Qiangguo Jin, Xianyao Zheng, Hui Cui et al.

Medical visual question answering (Med-VQA) is a crucial multimodal task in clinical decision support and telemedicine. Recent self-attention based methods struggle to effectively handle cross-modal semantic alignments between vision and language. Moreover, classification-based methods rely on predefined answer sets. Treating this task as a simple classification problem may make it unable to adapt to the diversity of free-form answers and overlook the detailed semantic information of free-form answers. In order to tackle these challenges, we introduce a Cross-Mamba Interaction based Multi-Task Learning (CMI-MTL) framework that learns cross-modal feature representations from images and texts. CMI-MTL comprises three key modules: fine-grained visual-text feature alignment (FVTA), cross-modal interleaved feature representation (CIFR), and free-form answer-enhanced multi-task learning (FFAE). FVTA extracts the most relevant regions in image-text pairs through fine-grained visual-text feature alignment. CIFR captures cross-modal sequential interactions via cross-modal interleaved feature representation. FFAE leverages auxiliary knowledge from open-ended questions through free-form answer-enhanced multi-task learning, improving the model's capability for open-ended Med-VQA. Experimental results show that CMI-MTL outperforms the existing state-of-the-art methods on three Med-VQA datasets: VQA-RAD, SLAKE, and OVQA. Furthermore, we conduct more interpretability experiments to prove the effectiveness. The code is publicly available at https://github.com/BioMedIA-repo/CMI-MTL.

Chao Pang, Yu Wang, Yi Jiang et al.

In this work, we propose MEDICO, a Multi-viEw Deep generative model for molecule generation, structural optimization, and the SARS-CoV-2 Inhibitor disCOvery. To the best of our knowledge, MEDICO is the first-of-this-kind graph generative model that can generate molecular graphs similar to the structure of targeted molecules, with a multi-view representation learning framework to sufficiently and adaptively learn comprehensive structural semantics from targeted molecular topology and geometry. We show that our MEDICO significantly outperforms the state-of-the-art methods in generating valid, unique, and novel molecules under benchmarking comparisons. In particular, we showcase the multi-view deep learning model enables us to generate not only the molecules structurally similar to the targeted molecules but also the molecules with desired chemical properties, demonstrating the strong capability of our model in exploring the chemical space deeply. Moreover, case study results on targeted molecule generation for the SARS-CoV-2 main protease (Mpro) show that by integrating molecule docking into our model as chemical priori, we successfully generate new small molecules with desired drug-like properties for the Mpro, potentially accelerating the de novo design of Covid-19 drugs. Further, we apply MEDICO to the structural optimization of three well-known Mpro inhibitors (N3, 11a, and GC376) and achieve ~88% improvement in their binding affinity to Mpro, demonstrating the application value of our model for the development of therapeutics for SARS-CoV-2 infection.

Linkuan Zhou, Yinghao Xia, Yufei Shen et al.

Unsupervised Domain Adaptation (UDA) is essential for deploying medical segmentation models across diverse clinical environments. Existing methods are fundamentally limited, suffering from semantically unaware feature alignment that results in poor distributional fidelity and from pseudo-label validation that disregards global anatomical constraints, thus failing to prevent the formation of globally implausible structures. To address these issues, we propose SHAPE (Structure-aware Hierarchical Unsupervised Domain Adaptation with Plausibility Evaluation), a framework that reframes adaptation towards global anatomical plausibility. Built on a DINOv3 foundation, its Hierarchical Feature Modulation (HFM) module first generates features with both high fidelity and class-awareness. This shifts the core challenge to robustly validating pseudo-labels. To augment conventional pixel-level validation, we introduce Hypergraph Plausibility Estimation (HPE), which leverages hypergraphs to assess the global anatomical plausibility that standard graphs cannot capture. This is complemented by Structural Anomaly Pruning (SAP) to purge remaining artifacts via cross-view stability. SHAPE significantly outperforms prior methods on cardiac and abdominal cross-modality benchmarks, achieving state-of-the-art average Dice scores of 90.08% (MRI->CT) and 78.51% (CT->MRI) on cardiac data, and 87.48% (MRI->CT) and 86.89% (CT->MRI) on abdominal data. The code is available at https://github.com/BioMedIA-repo/SHAPE.

Yu Wang, JingJie Zhang, Junru Jin et al.

Molecular representation learning (MRL) is a fundamental task for drug discovery. However, previous deep-learning (DL) methods focus excessively on learning robust inner-molecular representations by mask-dominated pretraining framework, neglecting abundant chemical reactivity molecular relationships that have been demonstrated as the determining factor for various molecular property prediction tasks. Here, we present MolCAP to promote MRL, a graph pretraining Transformer based on chemical reactivity (IMR) knowledge with prompted finetuning. Results show that MolCAP outperforms comparative methods based on traditional molecular pretraining framework, in 13 publicly available molecular datasets across a diversity of biomedical tasks. Prompted by MolCAP, even basic graph neural networks are capable of achieving surprising performance that outperforms previous models, indicating the promising prospect of applying reactivity information for MRL. In addition, manual designed molecular templets are potential to uncover the dataset bias. All in all, we expect our MolCAP to gain more chemical meaningful insights for the entire process of drug discovery.

Fengsheng Wang, Leyi Wei

With the rapid advances of image editing techniques in recent years, image manipulation detection has attracted considerable attention since the increasing security risks posed by tampered images. To address these challenges, a novel multi-scale multi-grained deep network (MSMG-Net) is proposed to automatically identify manipulated regions. In our MSMG-Net, a parallel multi-scale feature extraction structure is used to extract multi-scale features. Then the multi-grained feature learning is utilized to perceive object-level semantics relation of multi-scale features by introducing the shunted self-attention. To fuse multi-scale multi-grained features, global and local feature fusion block are designed for manipulated region segmentation by a bottom-up approach and multi-level feature aggregation block is designed for edge artifacts detection by a top-down approach. Thus, MSMG-Net can effectively perceive the object-level semantics and encode the edge artifact. Experimental results on five benchmark datasets justify the superior performance of the proposed method, outperforming state-of-the-art manipulation detection and localization methods. Extensive ablation experiments and feature visualization demonstrate the multi-scale multi-grained learning can present effective visual representations of manipulated regions. In addition, MSMG-Net shows better robustness when various post-processing methods further manipulate images.

Xianyao Zheng, Hong Yu, Hui Cui et al.

Medical visual question answering (Med-VQA) is a crucial multimodal task in clinical decision support and telemedicine. Recent methods fail to fully leverage domain-specific medical knowledge, making it difficult to accurately associate lesion features in medical images with key diagnostic criteria. Additionally, classification-based approaches typically rely on predefined answer sets. Treating Med-VQA as a simple classification problem limits its ability to adapt to the diversity of free-form answers and may overlook detailed semantic information in those answers. To address these challenges, we propose a knowledge graph enhanced cross-Mamba interaction (KG-CMI) framework, which consists of a fine-grained cross-modal feature alignment (FCFA) module, a knowledge graph embedding (KGE) module, a cross-modal interaction representation (CMIR) module, and a free-form answer enhanced multi-task learning (FAMT) module. The KG-CMI learns cross-modal feature representations for images and texts by effectively integrating professional medical knowledge through a graph, establishing associations between lesion features and disease knowledge. Moreover, FAMT leverages auxiliary knowledge from open-ended questions, improving the model's capability for open-ended Med-VQA. Experimental results demonstrate that KG-CMI outperforms existing state-of-the-art methods on three Med-VQA datasets, i.e., VQA-RAD, SLAKE, and OVQA. Additionally, we conduct interpretability experiments to further validate the framework's effectiveness.

Shuting Jin, Yi Jiang, Yimin Liu et al.

Background: Identifying new indications for approved drugs is a complex and time-consuming process that requires extensive knowledge of pharmacology, clinical data, and advanced computational methods. Recently, deep learning (DL) methods have shown their capability for the accurate prediction of drug repositioning. However, implementing DL-based modeling requires in-depth domain knowledge and proficient programming skills. Results: In this application, we introduce DeepDR, the first integrated platform that combines a variety of established DL-based models for disease- and target-specific drug repositioning tasks. DeepDR leverages invaluable experience to recommend candidate drugs, which covers more than 15 networks and a comprehensive knowledge graph that includes 5.9 million edges across 107 types of relationships connecting drugs, diseases, proteins/genes, pathways, and expression from six existing databases and a large scientific corpus of 24 million PubMed publications. Additionally, the recommended results include detailed descriptions of the recommended drugs and visualize key patterns with interpretability through a knowledge graph. Conclusion: DeepDR is free and open to all users without the requirement of registration. We believe it can provide an easy-to-use, systematic, highly accurate, and computationally automated platform for both experimental and computational scientists.

Qiangguo Jin, Hui Cui, Junbo Wang et al.

Semi-supervised learning (SSL) has attracted considerable attention in medical image processing. The latest SSL methods use a combination of consistency regularization and pseudo-labeling to achieve remarkable success. However, most existing SSL studies focus on segmenting large organs, neglecting the challenging scenarios where there are numerous tumors or tumors of small volume. Furthermore, the extensive capabilities of data augmentation strategies, particularly in the context of both labeled and unlabeled data, have yet to be thoroughly investigated. To tackle these challenges, we introduce a straightforward yet effective approach, termed iterative pseudo-labeling based adaptive copy-paste supervision (IPA-CP), for tumor segmentation in CT scans. IPA-CP incorporates a two-way uncertainty based adaptive augmentation mechanism, aiming to inject tumor uncertainties present in the mean teacher architecture into adaptive augmentation. Additionally, IPA-CP employs an iterative pseudo-label transition strategy to generate more robust and informative pseudo labels for the unlabeled samples. Extensive experiments on both in-house and public datasets show that our framework outperforms state-of-the-art SSL methods in medical image segmentation. Ablation study results demonstrate the effectiveness of our technical contributions.

Peng Zhou, Jianmin Wang, Chunyan Li et al.

While various models and computational tools have been proposed for structure and property analysis of molecules, generating molecules that conform to all desired structures and properties remains a challenge. Here, we introduce a multi-constraint molecular generation large language model, TSMMG, which, akin to a student, incorporates knowledge from various small models and tools, namely, the 'teachers'. To train TSMMG, we construct a large set of text-molecule pairs by extracting molecular knowledge from these 'teachers', enabling it to generate novel molecules that conform to the descriptions through various text prompts. We experimentally show that TSMMG remarkably performs in generating molecules meeting complex, natural language-described property requirements across two-, three-, and four-constraint tasks, with an average molecular validity of over 99% and success ratio of 82.58%, 68.03%, and 67.48%, respectively. The model also exhibits adaptability through zero-shot testing, creating molecules that satisfy combinations of properties that have not been encountered. It can comprehend text inputs with various language styles, extending beyond the confines of outlined prompts, as confirmed through empirical validation. Additionally, the knowledge distillation feature of TSMMG contributes to the continuous enhancement of small models, while the innovative approach to dataset construction effectively addresses the issues of data scarcity and quality, which positions TSMMG as a promising tool in the domains of drug discovery and materials science.

Linkuan Zhou, Zhexin Chen, Yufei Shen et al.

Automated segmentation of the fetal head in ultrasound images is critical for prenatal monitoring. However, achieving robust segmentation remains challenging due to the poor quality of ultrasound images and the lack of annotated data. Semi-supervised methods alleviate the lack of annotated data but struggle with the unique characteristics of fetal head ultrasound images, making it challenging to generate reliable pseudo-labels and enforce effective consistency regularization constraints. To address this issue, we propose a novel semi-supervised framework, ERSR, for fetal head ultrasound segmentation. Our framework consists of the dual-scoring adaptive filtering strategy, the ellipse-constrained pseudo-label refinement, and the symmetry-based multiple consistency regularization. The dual-scoring adaptive filtering strategy uses boundary consistency and contour regularity criteria to evaluate and filter teacher outputs. The ellipse-constrained pseudo-label refinement refines these filtered outputs by fitting least-squares ellipses, which strengthens pixels near the center of the fitted ellipse and suppresses noise simultaneously. The symmetry-based multiple consistency regularization enforces multi-level consistency across perturbed images, symmetric regions, and between original predictions and pseudo-labels, enabling the model to capture robust and stable shape representations. Our method achieves state-of-the-art performance on two benchmarks. On the HC18 dataset, it reaches Dice scores of 92.05% and 95.36% with 10% and 20% labeled data, respectively. On the PSFH dataset, the scores are 91.68% and 93.70% under the same settings.

Qiangguo Jin, Jiapeng Huang, Changming Sun et al.

The precise diagnosis of urinary stones is crucial for devising effective treatment strategies. The diagnostic process, however, is often complicated by the low contrast between stones and surrounding tissues, as well as the variability in stone locations across different patients. To address this issue, we propose a novel location embedding based pairwise distance learning network (LEPD-Net) that leverages low-dose abdominal X-ray imaging combined with location information for the fine-grained diagnosis of urinary stones. LEPD-Net enhances the representation of stone-related features through context-aware region enhancement, incorporates critical location knowledge via stone location embedding, and achieves recognition of fine-grained objects with our innovative fine-grained pairwise distance learning. Additionally, we have established an in-house dataset on urinary tract stones to demonstrate the effectiveness of our proposed approach. Comprehensive experiments conducted on this dataset reveal that our framework significantly surpasses existing state-of-the-art methods.

Qiangguo Jin, Hui Cui, Changming Sun et al.

Acquiring pixel-level annotations is often limited in applications such as histology studies that require domain expertise. Various semi-supervised learning approaches have been developed to work with limited ground truth annotations, such as the popular teacher-student models. However, hierarchical prediction uncertainty within the student model (intra-uncertainty) and image prediction uncertainty (inter-uncertainty) have not been fully utilized by existing methods. To address these issues, we first propose a novel inter- and intra-uncertainty regularization method to measure and constrain both inter- and intra-inconsistencies in the teacher-student architecture. We also propose a new two-stage network with pseudo-mask guided feature aggregation (PG-FANet) as the segmentation model. The two-stage structure complements with the uncertainty regularization strategy to avoid introducing extra modules in solving uncertainties and the aggregation mechanisms enable multi-scale and multi-stage feature integration. Comprehensive experimental results over the MoNuSeg and CRAG datasets show that our PG-FANet outperforms other state-of-the-art methods and our semi-supervised learning framework yields competitive performance with a limited amount of labeled data.

Qiangguo Jin, Hui Cui, Changming Sun et al.

The capability of generalization to unseen domains is crucial for deep learning models when considering real-world scenarios. However, current available medical image datasets, such as those for COVID-19 CT images, have large variations of infections and domain shift problems. To address this issue, we propose a prior knowledge driven domain adaptation and a dual-domain enhanced self-correction learning scheme. Based on the novel learning schemes, a domain adaptation based self-correction model (DASC-Net) is proposed for COVID-19 infection segmentation on CT images. DASC-Net consists of a novel attention and feature domain enhanced domain adaptation model (AFD-DA) to solve the domain shifts and a self-correction learning process to refine segmentation results. The innovations in AFD-DA include an image-level activation feature extractor with attention to lung abnormalities and a multi-level discrimination module for hierarchical feature domain alignment. The proposed self-correction learning process adaptively aggregates the learned model and corresponding pseudo labels for the propagation of aligned source and target domain information to alleviate the overfitting to noises caused by pseudo labels. Extensive experiments over three publicly available COVID-19 CT datasets demonstrate that DASC-Net consistently outperforms state-of-the-art segmentation, domain shift, and coronavirus infection segmentation methods. Ablation analysis further shows the effectiveness of the major components in our model. The DASC-Net enriches the theory of domain adaptation and self-correction learning in medical imaging and can be generalized to multi-site COVID-19 infection segmentation on CT images for clinical deployment.

Qiangguo Jin, Zhaopeng Meng, Changming Sun et al.

Automatic extraction of liver and tumor from CT volumes is a challenging task due to their heterogeneous and diffusive shapes. Recently, 2D and 3D deep convolutional neural networks have become popular in medical image segmentation tasks because of the utilization of large labeled datasets to learn hierarchical features. However, 3D networks have some drawbacks due to their high cost on computational resources. In this paper, we propose a 3D hybrid residual attention-aware segmentation method, named RA-UNet, to precisely extract the liver volume of interests (VOI) and segment tumors from the liver VOI. The proposed network has a basic architecture as a 3D U-Net which extracts contextual information combining low-level feature maps with high-level ones. Attention modules are stacked so that the attention-aware features change adaptively as the network goes "very deep" and this is made possible by residual learning. This is the first work that an attention residual mechanism is used to process medical volumetric images. We evaluated our framework on the public MICCAI 2017 Liver Tumor Segmentation dataset and the 3DIRCADb dataset. The results show that our architecture outperforms other state-of-the-art methods. We also extend our RA-UNet to brain tumor segmentation on the BraTS2018 and BraTS2017 datasets, and the results indicate that RA-UNet achieves good performance on a brain tumor segmentation task as well.

Qiangguo Jin, Zhaopeng Meng, Tuan D. Pham et al.

Automatic segmentation of retinal vessels in fundus images plays an important role in the diagnosis of some diseases such as diabetes and hypertension. In this paper, we propose Deformable U-Net (DUNet), which exploits the retinal vessels' local features with a U-shape architecture, in an end to end manner for retinal vessel segmentation. Inspired by the recently introduced deformable convolutional networks, we integrate the deformable convolution into the proposed network. The DUNet, with upsampling operators to increase the output resolution, is designed to extract context information and enable precise localization by combining low-level feature maps with high-level ones. Furthermore, DUNet captures the retinal vessels at various shapes and scales by adaptively adjusting the receptive fields according to vessels' scales and shapes. Three public datasets DRIVE, STARE and CHASE_DB1 are used to train and test our model. Detailed comparisons between the proposed network and the deformable neural network, U-Net are provided in our study. Results show that more detailed vessels are extracted by DUNet and it exhibits state-of-the-art performance for retinal vessel segmentation with a global accuracy of 0.9697/0.9722/0.9724 and AUC of 0.9856/0.9868/0.9863 on DRIVE, STARE and CHASE_DB1 respectively. Moreover, to show the generalization ability of the DUNet, we used another two retinal vessel data sets, one is named WIDE and the other is a synthetic data set with diverse styles, named SYNTHE, to qualitatively and quantitatively analyzed and compared with other methods. Results indicates that DUNet outperforms other state-of-the-arts.