Explore the vast range of titles available.

Signal-regulatory protein alpha (SIRPA) is a well-known inhibitor of phagocytosis when it complexes with CD47 expressed on target cells. Here we show that SIRPA decreased in vitro infection by a number of pathogenic viruses, including New World and Old World arenaviruses, Zika virus, vesicular stomatitis virus and pseudoviruses bearing the Machupo virus, Ebola virus and SARS-CoV-2 glycoproteins, but not HSV-1, MLV or mNoV. Moreover, mice with targeted mutation of the Sirpa gene that renders it non-functional were more susceptible to infection with the New World arenaviruses Junín virus vaccine strain Candid 1 and Tacaribe virus, but not MLV or mNoV. All SIRPA-inhibited viruses have in common the requirement for trafficking to a low pH endosomal compartment. This was clearly demonstrated with SARS-CoV-2 pseudovirus, which was only inhibited by SIRPA in cells in which it required trafficking to the endosome. Similar to its role in phagocytosis inhibition, SIRPA decreased virus internalization but not binding to cell surface receptors. We also found that increasing SIRPA levels via treatment with IL-4 led to even greater anti-viral activity. These data suggest that enhancing SIRPA’s activity could be a target for anti-viral therapies.

New world arenaviruses (NWAs) that cause viral hemorrhagic fever, such as Junín virus (JUNV), have few therapeutic options. Entry of these viruses into cells is mediated by binding to cell surface receptors, followed by endocytosis and trafficking to a low pH compartment. We showed previously that Signal Regulatory Protein Alpha (SIRPA), a critical cell surface receptor that inhibits macrophage phagocytic activity, also decreases internalization by NWAs and other RNA viruses that traffic to low pH compartments. Because SIRPA’s anti-phagocytic activity is thought to be the result of its ability to block integrin signaling needed for phagocytosis, we tested whether this was also the case for blockage of virus endocytosis. We show here that multiple proteins involved in the SIRPA/integrin signaling axis, including Src homology region 2 (SH2)-containing protein tyrosine phosphatase 2 (SHP2), src family kinases (SFK), particularly FYN, focal adhesion kinase (FAK), and alpha-integrin, play a role in viral endocytosis and in SIRPA’s ability to inhibit this process. This study thus reveals additional therapeutic targets for the inhibition of infection by viruses that traffic to acidic compartments.

Thermogravimetric (TG) experiments were carried out on the combustion of coal, peanut shells and their mixtures with different blending ratios in an air atmosphere to study the combustion characteristics of single particles and mixtures of the two, as well as the interactions between peanut shells and coal, and to find out the kinetic parameters of the combustion process, and to determine the optimal blending ratios. And the heat source plant 58 MW circulating fluidized bed boiler was used to carry out field mixing experiments on coal and peanut shells with different blending ratios to explore the pollutant emissions. The results show that blending peanut shell biomass can significantly improve the combustion characteristics of coal combustion and make combustion more likely to occur. The synergistic effect of peanut shells and coal was most significant when the blending ratio of peanut shells was 40%, and the mixture samples had the best combustion performance and the highest combined combustion properties index of 1.08 × 10–7. Compared with coal combustion alone, the emission concentrations of SO2 and NOx were reduced by 77.07% and 30.70% respectively when the ratio of peanut shell blending was 40%, the emission of dust showed a tendency of decreasing and then increasing, and the carbon content of fly ash was reduced from 7.55% to 5.86%. In our study, 40% peanut shells + 60% coal is the optimum blending ratio for mixed combustion. The results provide theoretical and technical support for energy saving, pollution reduction, and carbon emission reduction in coal-fired heating units.

Diffusion models have demonstrated their effectiveness across various generative tasks. However, when applied to medical image segmentation, these models encounter several challenges, including significant resource and time requirements. They also necessitate a multi-step reverse process and multiple samples to produce reliable predictions. To address these challenges, we introduce the first latent diffusion segmentation model, named SDSeg, built upon stable diffusion (SD). SDSeg incorporates a straightforward latent estimation strategy to facilitate a single-step reverse process and utilizes latent fusion concatenation to remove the necessity for multiple samples. Extensive experiments indicate that SDSeg surpasses existing state-of-the-art methods on five benchmark datasets featuring diverse imaging modalities. Remarkably, SDSeg is capable of generating stable predictions with a solitary reverse step and sample, epitomizing the model's stability as implied by its name. The code is available at https://github.com/lin-tianyu/Stable-Diffusion-Seg

Accurately detecting and localizing hallucinations is a critical task for ensuring high reliability of image captions. In the era of Multimodal Large Language Models (MLLMs), captions have evolved from brief sentences into comprehensive narratives, often spanning hundreds of words. This shift exponentially increases the challenge: models must now pinpoint specific erroneous spans or words within extensive contexts, rather than merely flag response-level inconsistencies. However, existing benchmarks lack the fine granularity and domain diversity required to evaluate this capability. To bridge this gap, we introduce DetailVerifyBench, a rigorous benchmark comprising 1,000 high-quality images across five distinct domains. With an average caption length of over 200 words and dense, token-level annotations of multiple hallucination types, it stands as the most challenging benchmark for precise hallucination localization in the field of long image captioning to date. Our benchmark is available at https://zyx-hhnkh.github.io/DetailVerifyBench/.

Hepatocellular Carcinoma diagnosis relies heavily on the interpretation of gigapixel Whole Slide Images. However, current computational approaches are constrained by fixed-resolution processing mechanisms and inefficient feature aggregation, which inevitably lead to either severe information loss or high feature redundancy. To address these challenges, we propose Hepato-LLaVA, a specialized Multi-modal Large Language Model designed for fine-grained hepatocellular pathology analysis. We introduce a novel Sparse Topo-Pack Attention mechanism that explicitly models 2D tissue topology. This mechanism effectively aggregates local diagnostic evidence into semantic summary tokens while preserving global context. Furthermore, to overcome the lack of multi-scale data, we present HepatoPathoVQA, a clinically grounded dataset comprising 33K hierarchically structured question-answer pairs validated by expert pathologists. Our experiments demonstrate that Hepato-LLaVA achieves state-of-the-art performance on HCC diagnosis and captioning tasks, significantly outperforming existing methods. Our code and implementation details are available at https://pris-cv.github.io/Hepto-LLaVA/.

Accurately grounding regions of interest (ROIs) is critical for diagnosis and treatment planning in medical imaging. While multimodal large language models (MLLMs) combine visual perception with natural language, current medical-grounding pipelines still rely on supervised fine-tuning with explicit spatial hints, making them ill-equipped to handle the implicit queries common in clinical practice. This work makes three core contributions. We first define Unified Medical Reasoning Grounding (UMRG), a novel vision-language task that demands clinical reasoning and pixel-level grounding. Second, we release U-MRG-14K, a dataset of 14K samples featuring pixel-level masks alongside implicit clinical queries and reasoning traces, spanning 10 modalities, 15 super-categories, and 108 specific categories. Finally, we introduce MedReasoner, a modular framework that distinctly separates reasoning from segmentation: an MLLM reasoner is optimized with reinforcement learning, while a frozen segmentation expert converts spatial prompts into masks, with alignment achieved through format and accuracy rewards. MedReasoner achieves state-of-the-art performance on U-MRG-14K and demonstrates strong generalization to unseen clinical queries, underscoring the significant promise of reinforcement learning for interpretable medical grounding.

Supervised Fine-Tuning (SFT) is the standard paradigm for domain adaptation, yet it frequently incurs the cost of catastrophic forgetting. In sharp contrast, on-policy Reinforcement Learning (RL) effectively preserves general capabilities. We investigate this discrepancy and identify a fundamental distributional gap: while RL aligns with the model's internal belief, SFT forces the model to fit external supervision. This mismatch often manifests as \"Confident Conflicts\" tokens characterized by low probability but low entropy. In these instances, the model is highly confident in its own prediction but is forced to learn a divergent ground truth, triggering destructive gradient updates. To address this, we propose Entropy-Adaptive Fine-Tuning (EAFT). Unlike methods relying solely on prediction probability, EAFT utilizes token-level entropy as a gating mechanism to distinguish between epistemic uncertainty and knowledge conflict. This allows the model to learn from uncertain samples while suppressing gradients on conflicting data. Extensive experiments on Qwen and GLM series (ranging from 4B to 32B parameters) across mathematical, medical, and agentic domains confirm our hypothesis. EAFT consistently matches the downstream performance of standard SFT while significantly mitigating the degradation of general capabilities.

Accurately grounding regions of interest (ROIs) is critical for diagnosis and treatment planning in medical imaging. While multimodal large language models (MLLMs) combine visual perception with natural language, current medical-grounding pipelines still rely on supervised fine-tuning with explicit spatial hints, making them ill-equipped to handle the implicit queries common in clinical practice. This work makes three core contributions. We first define Unified Medical Reasoning Grounding (UMRG), a novel vision-language task that demands clinical reasoning and pixel-level grounding. Second, we release U-MRG-14K, a dataset of 14K samples featuring pixel-level masks alongside implicit clinical queries and reasoning traces, spanning 10 modalities, 15 super-categories, and 108 specific categories. Finally, we introduce MedReasoner, a modular framework that distinctly separates reasoning from segmentation: an MLLM reasoner is optimized with reinforcement learning, while a frozen segmentation expert converts spatial prompts into masks, with alignment achieved through format and accuracy rewards. MedReasoner achieves state-of-the-art performance on U-MRG-14K and demonstrates strong generalization to unseen clinical queries, underscoring the significant promise of reinforcement learning for interpretable medical grounding.

The Segment Anything Model (SAM) has demonstrated strong and versatile segmentation capabilities, along with intuitive prompt-based interactions. However, customizing SAM for medical image segmentation requires massive amounts of pixel-level annotations and precise point- or box-based prompt designs. To address these challenges, we introduce PGP-SAM, a novel prototype-based few-shot tuning approach that uses limited samples to replace tedious manual prompts. Our key idea is to leverage inter- and intra-class prototypes to capture class-specific knowledge and relationships. We propose two main components: (1) a plug-and-play contextual modulation module that integrates multi-scale information, and (2) a class-guided cross-attention mechanism that fuses prototypes and features for automatic prompt generation. Experiments on a public multi-organ dataset and a private ventricle dataset demonstrate that PGP-SAM achieves superior mean Dice scores compared with existing prompt-free SAM variants, while using only 10\\% of the 2D slices.