The Segment Anything Model (SAM) has recently gained popularity in the field of image segmentation due to its impressive capabilities in various segmentation tasks and its prompt-based interface. However, recent studies and individual experiments have shown that SAM underperforms in medical image segmentation due to the lack of medical-specific knowledge. This raises the question of how to enhance SAM's segmentation capability for medical images. We propose the Medical SAM Adapter (Med-SA), which is one of the first methods to integrate SAM into medical image segmentation. Med-SA uses a light yet effective adaptation technique instead of fine-tuning the SAM model, incorporating domain-specific medical knowledge into the segmentation model. We also propose Space-Depth Transpose (SD-Trans) to adapt 2D SAM to 3D medical images and Hyper-Prompting Adapter (HyP-Adpt) to achieve prompt-conditioned adaptation. Comprehensive evaluation experiments on 17 medical image segmentation tasks across various modalities demonstrate the superior performance of Med-SA while updating only 2% of the SAM parameters (13M). Our code is released at https://github.com/KidsWithTokens/Medical-SAM-Adapter.

Developing reliable AI systems to assist human clinicians in multi-modal medical diagnosis has long been a key objective for researchers. Recently, Multi-modal Large Language Models (MLLMs) have gained significant attention and achieved success across various domains. With strong reasoning capabilities and the ability to perform diverse tasks based on user instructions, they hold great potential for enhancing medical diagnosis. However, directly applying MLLMs to the medical domain still presents challenges. They lack detailed perception of visual inputs, limiting their ability to perform quantitative image analysis, which is crucial for medical diagnostics. Additionally, MLLMs often exhibit hallucinations and inconsistencies in reasoning, whereas clinical diagnoses must adhere strictly to established criteria. To address these challenges, we propose MedAgent-Pro, an evidence-based reasoning agentic system designed to achieve reliable, explainable, and precise medical diagnoses. This is accomplished through a hierarchical workflow: at the task level, knowledge-based reasoning generate reliable diagnostic plans for specific diseases following retrieved clinical criteria. While at the case level, multiple tool agents process multi-modal inputs, analyze different indicators according to the plan, and provide a final diagnosis based on both quantitative and qualitative evidence. Comprehensive experiments on both 2D and 3D medical diagnosis tasks demonstrate the superiority and effectiveness of MedAgent-Pro, while case studies further highlight its reliability and interpretability. The code is available at https://github.com/jinlab-imvr/MedAgent-Pro.

Nuclei instance segmentation in histopathological images is crucial for biological analysis and cancer diagnosis. However, it faces two significant challenges: (1) poorly stained nuclei can lead to under-segmentation, as the background may be mistakenly identified as the foreground; and (2) deep textures within nuclei often result in fragmented instance predictions, as these textures can be misinterpreted as contours. To address these problems, this paper proposes a Structure Encoding and Interaction NEtwork, termed SEINE, which develops the nuclei structure modeling scheme and takes advantage of the similarity between nuclei structure to improve the integrality of instance segmentation. Specifically, SEINE introduces a contour-based structure encoding mechanism that integrates the correlation between nuclear structure and semantics, enabling a more accurate structural representation. Building on this encoding, we propose a structure-guided attention module, which uses clear nuclei as prototypes to guide the structural learning of unclear nuclei, thereby addressing the under-segmentation problem. Additionally, a position enhancement strategy applies a centroid distance constraint to reduce contour prediction errors, effectively mitigating fragmented instance segmentation. Extensive experiments demonstrate the effectiveness of SEINE, achieving state-of-the-art performance across four benchmark datasets. The code is available at https://github.com/zhangye-zoe/SEINE.

Weakly supervised segmentation methods have garnered considerable attention due to their potential to alleviate the need for labor-intensive pixel-level annotations during model training. Traditional weakly supervised nuclei segmentation approaches typically involve a two-stage process: pseudo-label generation followed by network training. The performance of these methods is highly dependent on the quality of the generated pseudo-labels, which can limit their effectiveness. In this paper, we propose a novel domain-adaptive weakly supervised nuclei segmentation framework that addresses the challenge of pseudolabel generation through cross-task interaction strategies. Specifically, our approach leverages weakly annotated data to train an auxiliary detection task, which facilitates domain adaptation of the segmentation network. To improve the efficiency of domain adaptation, we introduce a consistent feature constraint module that integrates prior knowledge from the source domain. Additionally, we develop methods for pseudo-label optimization and interactive training to enhance domain transfer capabilities. We validate the effectiveness of our proposed method through extensive comparative and ablation experiments conducted on six datasets. The results demonstrate that our approach outperforms existing weakly supervised methods and achieves performance comparable to or exceeding that of fully supervised methods. Our code is available at https://github.com/zhangye-zoe/DAWN.

While Vision-Language Models (VLMs) have achieved notable progress in computational pathology (CPath), the gigapixel scale and spatial heterogeneity of Whole Slide Images (WSIs) continue to pose challenges for multimodal understanding. Existing alignment methods struggle to capture fine-grained correspondences between textual descriptions and visual cues across thousands of patches from a slide, compromising their performance on downstream tasks. In this paper, we propose PathFLIP (Pathology Fine-grained Language-Image Pretraining), a novel framework for holistic WSI interpretation. PathFLIP decomposes slide-level captions into region-level subcaptions and generates text-conditioned region embeddings to facilitate precise visual-language grounding. By harnessing Large Language Models (LLMs), PathFLIP can seamlessly follow diverse clinical instructions and adapt to varied diagnostic contexts. Furthermore, it exhibits versatile capabilities across multiple paradigms, efficiently handling slide-level classification and retrieval, fine-grained lesion localization, and instruction following. Extensive experiments demonstrate that PathFLIP outperforms existing large-scale pathological VLMs on four representative benchmarks while requiring significantly less training data, paving the way for fine-grained, instruction-aware WSI interpretation in clinical practice.