Early Access

Unsupervised anomaly detection (UAD) in brain MRI is crucial for early diagnosis, yet generalizing existing methods across diverse diseases, sequences, and missing data scenarios remains a significant challenge. Current reconstruction-based methods often fail to detect subtle anomalies, while conventional translation methods lack flexibility regarding input sequences. To address these limitations,...

中文摘要翻译待生成

Reconstructing dynamic surgical scenes from endoscopic videos remains a fundamental challenge in robot-assisted surgery. Existing methods primarily focus on deformable tissues, overlooking the presence of articulated instruments. To bridge this gap, we present EndoLRMGS, the first unified framework capable of reconstructing both deformable tissue and articulated instruments in a modular approach f...

中文摘要翻译待生成

Myocardial echocardiography segmentation is useful for cardiac function assessment, yet remains challenging due to continuous myocardial motion and deformation during the cardiac cycle, as well as boundary ambiguity caused by speckle noise. Existing memory-based segmentation methods typically construct memory banks by accumulating single-frame representations, where the stored memory items fail to...

中文摘要翻译待生成

Histopathological analysis of stained tissue remains central to biomedical research and clinical care. Virtual staining (VS) offers a promising alternative, with potential to reduce costs and streamline workflows, yet hallucinations pose serious risks to clinical reliability. Here, we formalize the problem of hallucination detection in VS and propose a scalable post-hoc baseline method: Neural Hal...

中文摘要翻译待生成

Ordinal regression is well-known for lever-aging the underlying inherent order between successive categories to obtain additional regularization beyond traditional probabilistic classification mechanism. However, there are challenges in real-world medical grading tasks: 1) The uneven distribution of disease severity levels, characterized by a long-tailed format, complicates the ordinal regression ...

中文摘要翻译待生成

Array topology in magnetocardiography system dictates spatial resolution and clinical applicability, impacting topographic reconstruction, which affects the accuracy of feature and source localization. Existing array designs rely on task-driven or empirical criteria without accounting for intrinsic imaging properties to ensure complete high-precision reconstruction and broad applicability, and ana...

中文摘要翻译待生成

Limited-angle cardiac CT reconstruction is a severely ill-posed problem, where incomplete angular coverage leads to strong artifacts and structural distortions. Although diffusion-based methods have shown strong potential for improving reconstruction quality, their high computational cost, large memory demand, and slow inference remain major barriers to practical clinical deployment. To address th...

中文摘要翻译待生成

Mild cognitive impairment (MCI) is the prodromal stage of dementia involving complex interactions between the brain and peripheral organs. Emerging evidence indicates that heart dysfunction and gut microbiota dysbiosis can contribute to MCI pathogenesis. Yet, these discoveries of cross-organ interactions have not been applied to assist MCI diagnosis. In this work, we propose a novel diagnostic fra...

中文摘要翻译待生成

Accurate correspondence matching in coronary angiography images is crucial for reconstructing 3D coronary artery structures, which is essential for precise diagnosis and treatment planning of coronary artery disease (CAD). Traditional matching methods for natural images often fail to generalize to X-ray images due to inherent differences such as lack of texture, lower contrast, and overlapping structures, compounded by insufficient training data. To address these challenges, we propose a novel pipeline that generates realistic paired coronary angiography images using a diffusion model conditioned on 2D projections of 3D reconstructed meshes from Coronary Computed Tomography Angiography (CCTA), providing high-quality synthetic data for training. Additionally, we employ large-scale image foundation models to guide feature aggregation, enhancing correspondence matching accuracy by focusing on semantically relevant regions and keypoints. Our approach demonstrates superior matching performance on synthetic datasets and effectively generalizes to real-world datasets, offering a practical solution for this task. Furthermore, our work investigates the efficacy of different foundation models in correspondence matching, providing novel insights into leveraging advanced image foundation models for medical imaging applications.

在冠状动脉造影图像中，准确的对应匹配对于重建3D冠状动脉结构至关重要，这对于冠状动脉疾病（CAD）的精确诊断和治疗规划是必不可少的。传统的自然图像匹配方法由于缺乏纹理、对比度较低以及结构重叠等固有差异，往往无法泛化到X射线图像上。

Deep neural networks for medical image segmentation are often overconfident, compromising both reliability and clinical utility. In this work, we propose differentiable formulations of marginal L1 Average Calibration Error (mL1-ACE) as an auxiliary loss that can be computed on a per-image basis. We compare both hard-and soft-binning approaches to directly improve pixel-wise calibration. Our experiments on four datasets (ACDC, AMOS, KiTS, BraTS) demonstrate that incorporating mL1-ACE significantly reduces calibration errors, particularly Average Calibration Error (ACE) and Maximum Calibration Error (MCE), while largely maintaining high Dice Similarity Coefficients (DSCs). We find that the soft-binned variant yields the greatest improvements in calibration, over the DSC plus cross-entropy loss baseline, but often compromises segmentation performance, with hard-binned mL1-ACE maintaining segmentation performance, albeit with weaker calibration improvement. To gain further insight into calibration performance and its variability across an imaging dataset, we introduce dataset reliability histograms, an aggregation of per-image reliability diagrams. The resulting analysis highlights improved alignment between predicted confidences and true accuracies. Overall, our approach provides practitioners with explicit control over the calibration-accuracy trade-off, enabling more reliable integration of deep learning methods into clinical workflows. We share our code here: https://github.com/ cai4cai/Average-Calibration-Losses.

用于医学图像分割的深度神经网络往往过于自信，损害了可靠性和临床实用性。在这项工作中，我们提出了边缘L1平均校准误差（mL1-ACE）的可微形式作为辅助损失，可以在每幅图像上计算。我们比较了硬分箱和软分箱方法以直接改进逐像素校准。在四个数据集（ACDC、AMOS、KiTS、BraTS）上的实验表明，加入mL1-ACE显著降低了校准误差，特别是平均校准误差（ACE）和最大校准误差（MCE），同时大部分保持了高Dice相似系数（DSC）。我们发现，软分箱变体在DSC加交叉熵损失基线上取得了最大的校准改进，但往往牺牲了分割性能，而硬分箱mL1-ACE保持了分割性能，但校准改进较弱。为了进一步洞察校准性能及其在影像数据集上的变异性，我们引入了数据集可靠性直方图，即每幅图像可靠性图的聚合。由此分析突出了预测置信度与真实准确度之间对齐的改善。总体而言，我们的方法使从业者能够显式控制校准-准确度权衡，从而将深度学习方法更可靠地集成到临床工作流程中。我们在https://github.com/cai4cai/Average-Calibration-Losses分享代码。

Sampling jitter, i.e., random deviations in the time instants when samples are taken, causes frequency-dependent noise that reduces the signal-to-noise ratio (SNR). This paper generalizes the concept of jitter to magnetoencephalography (MEG) sensor arrays that spatially sample the quasistatic magnetic field due to brain activity. It is shown that spatial jitter, i.e., random deviations in the MEG ...

中文摘要翻译待生成

Age-related macular degeneration with abnormal blood vessel growth (neovascular AMD) is the leading cause of vision loss in elderly populations. While anti-VEGF injections are the standard treatment, they present financial burdens for patients and vary in effectiveness. Predicting treatment efficacy is therefore crucial for patient care. Current prediction methods fail to fully integrate informati...

中文摘要翻译待生成

Organoids are three-dimensional (3D) in vitro models for studying tissue development, disease progression, and physiological responses. Holotomography (HT) enables long-term, label-free imaging of live organoids by reconstructing volumetric refractive-index (RI) maps, but quantitative analysis is limited by the missing-cone artifact, which introduces anisotropic resolution and axial distortion. Here, we present a quantitative analysis framework that addresses the missing-cone problem at the level of image representation rather than reconstruction. We introduce morphology-preserving holotomography (MP-HT), a torus-shaped spatial filtering strategy that emphasizes high-spatial-frequency RI texture while suppressing low-frequency components most susceptible to missing-cone-induced distortion. Based on MP-HT, we develop a 3D segmentation pipeline for robust separation of epithelial and luminal structures, together with a model-based RI quantification approach that incorporates the system point spread function to enable morphology-independent estimation of dry-mass density and total dry mass. We apply the framework to long-term imaging of live hepatic organoids undergoing expansion, collapse, and fusion. In representative organoids, the framework provides consistent segmentation across diverse geometries and enables quantitative characterization of epithelial-lumen remodeling, collapse-associated loss of morphometric stability, and transient biophysical fluctuations during fusion. Overall, this work establishes a physically transparent and reproducible approach for quantitative, label-free analysis of organoid dynamics in 3D.

类器官是用于研究组织发育、疾病进展和生理反应的三维体外模型。全息断层成像（HT）通过重建体积折射率（RI）图，能够对活体类器官进行长期、无标记成像，但定量分析受到缺失锥伪影的限制，该伪影引入了各向异性分辨率和轴向畸变。他...

Motion remains a problem in clinical MRI, largely because all existing effective correction methods come with a penalty – constraints on pulse sequence parameters, expensive/bulky equipment, or extra steps in the workflow. The Pilot Tone (PT) is a small device that does not physically contact with the patient and only minimally impacts workflows. However, its signals can be difficult to process du...

中文摘要翻译待生成

Low-dose computed tomography (CT) denoising is crucial for reduced radiation exposure while ensuring diagnostically acceptable image quality. Despite significant advancements driven by deep learning (DL) in recent years, existing DL-based methods, typically trained on a specific dose level and anatomical region, struggle to handle diverse noise characteristics and anatomical heterogeneity during varied scanning conditions, limiting their generalizability and robustness in clinical scenarios. In this paper, we propose FoundDiff, a foundational diffusion model for unified and generalizable LDCT denoising across various dose levels and anatomical regions. FoundDiff employs a two-stage strategy: (i) dose-anatomy perception and (ii) adaptive denoising. First, we develop a dose- and anatomy-aware contrastive language-image pre-training model (DA-CLIP) to achieve robust dose and anatomy perception by leveraging specialized contrastive learning strategies to learn continuous representations that quantify ordinal dose variations and identify salient anatomical regions. Second, we design a dose-and anatomy-aware diffusion model (DA-Diff) to perform adaptive and generalizable denoising by synergistically integrating the learned dose and anatomy embeddings from DA-CLIP into diffusion process via a novel dose and anatomy conditional block (DACB) based on Mamba. Extensive experiments on a large simulated multi-dose CT dataset spanning three anatomical regions, together with cross-dataset evaluations on Mayo-2016, CQ500, and piglet datasets, demonstrate superior denoising performance and strong generalization to unseen dose levels and anatomical regions. The codes and models are available at https: //github.com/hao1635/FoundDiff.

低剂量计算机断层扫描（CT）去噪对于减少辐射暴露同时确保诊断可接受的图像质量至关重要。尽管近年来深度学习（DL）推动了显著进展，但现有的基于DL的方法通常针对特定剂量水平和解剖区域进行训练，难以处理在...期间出现的多样噪声特性和解剖异质性。

Limited-angle computed tomography (LACT) improves temporal resolution and reduces radiation dose, but suffers from severe artifacts due to missing projections. Clinical workflows record abundant patient- and acquisition-level metadata, yet such information remains underutilized in image reconstruction. To tackle the ill-posed LACT inverse problem, we propose a metadata-guided two-stage diffusion f...

中文摘要翻译待生成

Fiber tract segmentation is crucial for clinical applications such as brain function interpretation and surgical planning. Existing methods typically adopt either a cortical-parcellation-based or fiber clustering approach, but fail to simultaneously integrate heterogeneous information (e.g., streamline shape, point position, anatomical priors). In this work, we propose Fiber HGNN, a novel heteroge...

中文摘要翻译待生成

Vision-Language Models (VLMs) in visual question answering (VQA) offer a unique opportunity to enhance intra-operative decision-making, promote intuitive interactions, and significantly advance surgical education. However, the development of VLMs for surgical VQA is challenging due to limited datasets and the risk of overfitting and catastrophic forgetting during full fine-tuning of pretrained weights. While parameter-efficient techniques like Low-Rank Adaptation (LoRA) and Matrix of Rank Adaptation (MoRA) address adaptation challenges, their uniform parameter distribution overlooks the feature hierarchy in deep networks, where earlier layers, that learn general features, require more parameters than later ones. This work introduces PitVQA++ with an Open-ended PitVQA dataset and vector matrix-low-rank adaptation (Vector-MoLoRA), an innovative VLM fine-tuning approach for adapting GPT-2 to pituitary surgery. Open-Ended PitVQA comprises 109,173 frames from 25 procedural videos with 795,270 question-answer sentence pairs, covering key surgical elements such as phase and step recognition, context understanding, tool detection, localization, and interactions recognition. Vector-MoLoRA incorporates the principles of LoRA and MoRA to develop a matrix-low-rank adaptation strategy that employs rank vectors to allocate more parameters to earlier layers, gradually reducing them in the later layers. Our approach, validated on the Open-Ended PitVQA and EndoVis18-VQA datasets, effectively mitigates catastrophic forgetting while significantly enhancing performance over recent baselines. Performance-rejection analysis further highlights Vector-MoLoRA’s enhanced reliability and trust-worthiness in handling uncertain predictions. Our source code and dataset is available at https://github.com/ HRL-Mike/PitVQA-Plus.

视觉语言模型在视觉问答中为增强术中决策、促进直观交互以及显著推进外科教育提供了独特的机会。然而，由于数据集有限，以及在预训练权重完全微调过程中存在过拟合和灾难性遗忘的风险，用于手术视觉问答的视觉语言模型的开发面临挑战。

Developing a reliable EEG-based Brain Computer Interface (BCI) system typically requires large and diverse training datasets, but collecting sufficient data remains challenging due to subject fatigue and interindividual variability. To address these limitations, this study proposes a reinforcement learning-enhanced EEG diffusion (RLED) framework for adaptive data augmentation in endogenous EEG tasks, with a focus on motor imagery and emotion recognition. The framework integrates a reinforcement learning mechanism to dynamically regulate the diffusion training process and achieve a flexible balance among temporal, spectral, and class-related features. Experiments on four datasets demonstrate that the proposed method generates high-quality synthetic EEG signals and consistently improves classification performance. These findings show that the proposed RLED framework may serve as a promising tool for EEG data augmentation and generalization in practical BCI applications.

开发一个可靠的基于脑电图的脑机接口系统通常需要大量多样的训练数据集，但受试者疲劳和个体差异使得收集足够数据仍然具有挑战性。为了解决这些限制，本研究提出了一种强化学习增强的脑电图扩散框架，用于内源性脑电图任务中的自适应数据增强。

Intraoperative registration of real-time ultra-sound (iUS) to preoperative Magnetic Resonance Imaging (MRI) remains an unsolved problem due to severe modality-specific differences in appearance, resolution, and field-of-view. To address this, we propose a novel 3D cross-modal keypoint descriptor for MRI–iUS matching and registration. Our approach employs a patient-specific matching-by-synthesis approach, generating synthetic iUS volumes from preoperative MRI. This enables supervised contrastive training to learn a shared descriptor space. A probabilistic keypoint detection strategy is then employed to identify anatomically salient and modality-consistent locations. During training, a curriculum-based triplet loss with dynamic hard negative mining is used to learn descriptors that are i) robust to iUS artifacts such as speckle noise and limited coverage, and ii) rotation-invariant. At inference, the method detects keypoints in MR and real iUS images and identifies sparse matches, which are then used to perform rigid registration. Our approach is evaluated using 3D MRI-iUS pairs from the ReMIND dataset. Experiments show that our approach outperforms state-of-the-art keypoint matching methods across 11 patients, with an average precision of 69.8%. For image registration, our method achieves a competitive mean Target Registration Error of 2.39 mm on the ReMIND2Reg benchmark. Compared to existing iUS-MR registration approaches, our framework is interpretable, requires no manual initialization, and shows robustness to iUS field-of-view variation. Code, data and model weights are available at https://github.com/morozovdd/CrossKEY .

术中实时超声（iUS）与术前磁共振成像（MRI）的配准由于外观、分辨率和视野上的严重模态特异性差异仍然是一个未解决的问题。为了解决这个问题，我们提出了一种新的用于MRI-iUS匹配和配准的三维跨模态关键点描述符。我们的方法采用了一种特定患者的合成匹配方法...

Two-dimensional cell culture models have long been a cornerstone of biomedical research; however, they often fail to accurately replicate the in vivo environment. In recent years, three-dimensional (3D) cell cultures, particularly 3D spheroid models, have gained recognition for their ability to better mimic the complexities of the in vivo environment, making them valuable tools for studying cellular behavior and responses. Tumor spheroids, in particular, have significant applications in anticancer therapy evaluation, providing a more physiologically relevant model by simulating the spatial architecture and microenvironment of tumors. However, due to the limitations imposed by optical diffraction and background noise in 3D imaging, traditional imaging methods are unable to accurately resolve the growth, morphological changes, and drug responses of tumor spheroids. To address this issue, super-resolution imaging technologies have emerged. Structured illumination microscopy (SIM) combined with reconstruction algorithms can effectively enhance resolution, but challenges such as limited light penetration of single-photon imaging and high background noise remain in 3D imaging. In this paper, an advanced SIM technology with large depth and low noise 3D imaging capability is developed. This study introduces a novel frequency-specific denoising method (FSDM) to effectively reduce noise through adjusting the weights of high-frequency signals to preserve image details. The FSDM optimization significantly reduces background interference from deeper tissue layers, improving image details and the overall quality of 3D imaging. For the first time, scanning SIM is integrated with two-photon microscopy (TPEF-SIM) for 3D imaging, leveraging the strengths of both techniques to enhance resolution and overcome light penetration limitations.

二维细胞培养模型长期以来是生物医学研究的基石；然而，它们常常无法精确再现体内环境。近年来，三维（3D）细胞培养，特别是3D球体模型，因能更好地模拟体内环境的复杂性而获得认可，使其成为研究细胞...的有价值工具。

Functional MRI (fMRI) and single-cell transcri ptomics are pivotal in Alzheimer’s disease (AD) research, each providing unique insights into neural function and molecular mechanisms. However, integrating these complementary modalities remains largely unexplored. Here, we introduce scBIT, a novel method for enhancing AD prediction by combining fMRI with single-nucleus RNA (snRNA). scBIT leverages snRNA as an auxiliary modality, significantly improving fMRI-based prediction models and providing comprehensive interpretability. It employs a sampling strategy to segment snRNA data into cell-type-specific gene networks and utilizes a self-explainable graph neural network to extract critical subgraphs. Additionally, we use demographic and genetic similarities to pair snRNA and fMRI data across individuals, enabling robust cross-modal learning. Extensive experiments validate scBIT’s effectiveness in revealing intricate brain region-gene associations and enhancing diagnostic prediction accuracy. By advancing brain imaging transcriptomics to the single-cell level, scBIT sheds new light on biomarker discovery in AD research. Experimental results show that incorporating snRNA data into the scBIT model significantly boosts accuracy, improving binary classification by 3.39% and five-class classification by 26.59%. The codes were implemented in Python and have been released on GitHub (https://github.com/77YQ77/scBIT) and Zenodo (https://zenodo.org/records/11599030) with detailed instructions.

中文摘要翻译待生成

4D Flow Magnetic Resonance Imaging (4D Flow MRI) is a non-invasive technique for volumetric, time-resolved blood flow quantification. However, apparent trade-offs between acquisition time, image noise, and resolution limit clinical applicability. In particular, in regions of highly transient flow, coarse temporal resolution can hinder accurate capture of physiologically relevant flow variations. Deep learning-based post-processing techniques have shown promise in overcoming these issues using so-called super-resolution networks. However, while existing super-resolution research has primarily focused on spatial upsampling, temporal super-resolution remains largely unexplored. The aim of this study was therefore to implement and evaluate a residual data-driven network for temporal super-resolution 4D Flow MRI. To achieve this, an existing spatial network (4DFlowNet) was re-designed for temporal upsampling, adapting input dimensions, and optimizing internal layer structures. The model was trained and tested on synthetic 4D Flow MRI data derived from patient-specific in-silico models, followed by additional evaluation on clinically acquired in-vivo datasets. Overall, excellent performance was achieved with input velocities effectively denoised and temporally upsampled, with a mean absolute error (MAE) of 1.0 cm/s in an unseen in-silico setting, outperforming deterministic alternatives (linear interpolation MAE = 2.3 cm/s, sinc interpolation MAE = 2.6 cm/s). Further, the network synthesized high-resolution temporal information from unseen low-resolution in-vivo data, with strong correlation observed at peak flow frames. As such, our results highlight the potential of utilizing data-driven neural networks for temporal super-resolution 4D Flow MRI, enabling high-frame-rate flow quantification without extending acquisition times beyond clinically acceptable limits.

中文摘要翻译待生成

4D (3D over time) fetal heart reconstruction improves detection and functional assessment of congenital malformations compared with 2D methods, but remains challenging due to the lack of publicly available 4D echocardiography datasets, the burden of full 3D/4D annotations, and the computational cost of volumetric networks. To address these challenges, we introduce a 2.5D radial-slicing paradigm th...

中文摘要翻译待生成

Glaucoma is a leading cause of irreversible blindness worldwide, with asymptomatic early stages often delaying diagnosis and treatment. Early and accurate diagnosis requires integrating complementary information from multiple ocular imaging modalities. However, most existing studies rely on single- or dual-modality imaging, such as fundus and optical coherence tomography (OCT), for coarse binary classification, thereby restricting the exploitation of complementary information and hindering both early diagnosis and stage-specific treatment. To address these limitations, we propose glaucoma lesion evaluation and analysis with multimodal imaging (GLEAM), the first publicly available tri-modal glaucoma dataset comprising scanning laser ophthalmoscopy fundus images, circumpapillary OCT images, and visual field pattern deviation maps, annotated with four disease stages, enabling effective exploitation of multimodal complementary information and facilitating accurate diagnosis and treatment across disease stages. To effectively integrate cross-modal information, we propose hierarchical attentive masked modeling (HAMM) for multimodal glaucoma classification. Our framework employs hierarchical attentive encoders and light decoders to focus cross-modal representation learning on the encoder. The attention module, named multimodal-channel graph attention (MCGA), boosts glaucoma classification performance by emulating two key clinical reasoning steps: first, it uses a multi-head modality gating mechanism to replicate ophthalmologists’ confidence scoring of fundus, OCT, and VF modalities; then, MCGA leverages a relational graph attention network to cross-examine structural-functional consistencies of weighted modalities. The experiments on GLEAM demonstrate that tri-modal fusion significantly outperforms single-modal and dual-modal configurations. Moreover, our proposed HAMM achieves superior performance compared with state-of-the-art multimodal learning methods. The dataset and code are publicly available via https://github.com/microewing/HAMM.

青光眼是全球不可逆失明的主要原因，其无症状早期阶段常常延误诊断和治疗。早期准确诊断需要整合来自多种眼部成像模式的互补信息。然而，大多数现有研究依赖于单模态或双模态成像，如眼底和光学相干断层扫描（OCT），用于粗略的二元分类...