MICCAI 2025 Papers — Page 3
International Conference on Medical Image Computing and Computer-Assisted Intervention · 1027 papers
CurConMix: A Curriculum Contrastive Learning Framework for Enhancing Surgical Action Triplet Recognition
Jeon, Yongjun (Sungkyunkwan University), Jung, Kyu-Hwan (Sungkyunkwan University)
RecognitionRepresentation LearningTransformerSupervised Fine-TuningContrastive LearningVideo
🎯 What it does: The study focuses on recognizing action triplets (instrument, verb, target) in surgical videos and proposes a framework called CurConMix.
CXR-CML: Improved zero-shot classification of long-tailed multi-label diseases in Chest X-Rays
Madhipati, Rajesh (Friedrich-Alexander University Erlangen-Nuremberg), Maier, Andreas (Friedrich-Alexander-Universität Erlangen)
ClassificationContrastive LearningImageBiomedical Data
🎯 What it does: This paper proposes an improved zero-shot multi-label long-tail disease classification method called CXR-CML based on CLIP.
CXR-TFT: Multi-Modal Temporal Fusion Transformer for Predicting Chest X-ray Trajectories
Arora, Mehak (Duke University), Kamaleswaran, Rishikesan (Duke University)
Anomaly DetectionTransformerVision Language ModelImageMultimodalityTime SeriesBiomedical Data
🎯 What it does: This study proposes CXR-TFT, a multimodal temporal fusion Transformer, to predict the future imaging features and pathological changes of chest X-rays in ICU patients.
Cycle Context Verification for In-Context Medical Image Segmentation
Hu, Shishuai (Northwestern Polytechnical University), Xia, Yong (Northwestern Polytechnical University)
SegmentationImageBiomedical DataUltrasound
🎯 What it does: This paper proposes a Circular Context Validation (CCV) framework that utilizes a self-check mechanism during the inference phase and query-specific prompts to enhance the performance of context-based learning in medical image segmentation.
CytoSAE: Interpretable Cell Embeddings for Hematology
Dasdelen, Muhammed Furkan, Schneider, Steffen (Institute of Computational Biology, Helmholtz Munich)
Explainability and InterpretabilityRepresentation LearningAuto EncoderImageBiomedical Data
🎯 What it does: Using sparse autoencoders for unsupervised decomposition of blood cell embeddings to learn interpretable morphological concepts.
D-CAM: Learning Generalizable Weakly-Supervised Medical Image Segmentation from Domain-invariant CAM
Yi, Jingjun (Tencent Jarvis Lab), Huang, Feiyue (Guangxi Medical University)
SegmentationDomain AdaptationConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: A D-CAM method based on frequency domain decoupling is proposed, which utilizes weakly supervised image-level labels to train a classification network in the source domain to generate domain-invariant class activation maps (CAM), and then uses these pseudo-labels to train a segmentation model in the target domain, achieving generalizable medical image segmentation.
D2Diff: A Dual-Domain Diffusion Model for Accurate Multi-Contrast MRI Synthesis
Dayarathna, Sanuwani (Monash University), Chen, Zhaolin (Monash University)
SegmentationGenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A dual-domain diffusion model D2Diff is proposed for synthesizing missing or defective contrast images from existing multi-contrast MRI.
D2MAE: Diffusional Deblurring MAE for Ultrasound Image Pre-training
Kang, Qingbo, Lao, Qicheng (Ningbo Fregty Optoelectronics Technology Co., Ltd)
RestorationRepresentation LearningTransformerDiffusion modelAuto EncoderImageBiomedical DataUltrasound
🎯 What it does: This paper proposes D MAE, which integrates the diffusion deblurring process with MAE for self-supervised pre-training of ultrasound images.
D3M: Deformation-Driven Diffusion Model for Synthesis of Contrast-Enhanced MRI with Brain Tumors
Pang, Haowen (Beijing Institute of Technology), Ye, Chuyang (Beijing Institute of Technology)
RestorationGenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A D-M model is proposed to embed spatial deformation in diffusion models for synthesizing enhanced T1 images of brain tumors from non-enhanced MRI.
D4Recon: Dual-stage Deformation and Dual-scale Depth Guidance for Endoscopic Reconstruction
Basak, Hritam (Stony Brook University), Yin, Zhaozheng (Stony Brook University)
RestorationDepth EstimationGaussian SplattingImageVideo
🎯 What it does: This paper presents the D4 Recon framework, which achieves real-time high-fidelity endoscopic reconstruction by combining dual-stage spatial and temporal deformation modeling with dual-scale depth guidance, utilizing dynamic 3D Gaussian splatting for fine geometry and texture reconstruction.
DC-Seg: Disentangled Contrastive Learning for Brain Tumor Segmentation with Missing Modalities
Li, Haitao (Zhejiang University), Huang, Zhengxing (First Affiliated Hospital of Zhejiang University School of Medicine)
SegmentationContrastive LearningMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A model named DC-Seg is proposed for brain tumor segmentation, which maintains robustness in the absence of multimodal data.
DCKAN: A Dual-Coordinate KAN Framework for Fibrous Cap Segmentation on Carotid OCT
Wan, Tonghua (Huazhong University of Science and Technology), Qiu, Wu (Huazhong University of Science and Technology)
SegmentationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A dual-coordinate (Cartesian and Polar) KAN framework is designed and implemented for the segmentation of fibrous caps in carotid OCT images.
DCT-Net: Dual-branch CT Reconstruction from Orthogonal X-rays with Diffusion Model and Contrastive Learning
Zhang, Zhiyu (Tianjin University of Technology), Liao, Zhijun (Shenzhen Institute of Advanced Technology)
RestorationGenerationDiffusion modelGenerative Adversarial NetworkContrastive LearningImageComputed Tomography
🎯 What it does: A dual-branch CT reconstruction network DCT-Net is proposed, which utilizes a conditional diffusion model to generate bone suppression X-ray images. The original and enhanced X-rays are input into a dual generator GAN, achieving 3D CT reconstruction from frontal and lateral X-rays through perceptual loss and multi-path loss.
Decentralized Noise Handling in Medical Imaging: Encoder-Decoder Based Federated Imputation for Robust Training
Chang, Yunyoung (Gachon University), Noh, Wonjong (Hallym University)
ClassificationRestorationFederated LearningTransformerAuto EncoderImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An end-to-end encoding-decoding federated completion framework is proposed, which first performs noise prediction and completion on medical images in a federated learning environment, followed by classification training.
Deep Association Multimodal Learning for Zero-shot Spatial Transcriptomics Prediction
Zhou, Yijing (East China Normal University), Wang, Yan (East China Normal University)
Graph Neural NetworkTransformerLarge Language ModelPrompt EngineeringImageMultimodalityBiomedical Data
🎯 What it does: This paper proposes a deep associative multimodal framework that can predict gene expression of spatial transcriptomics from pathological images under zero-shot conditions.
Deep Knowledge-Infused Transformer for NSCLC Lymph Node Station Metastasis Prediction: Development of an AI-Powered Intraoperative Decision System
Zhao, Jie (Peking University), Dong, Bin (Peking University)
Graph Neural NetworkTransformerBiomedical Data
🎯 What it does: This study investigates a Deep Knowledge Injection Transformer model (DKiT) for predicting lymph node station metastasis in patients with non-small cell lung cancer, and based on this, a real-time intraoperative decision support system was developed.
Deep Learning Framework for Managing Inter-Reader Variability in Background Parenchymal Enhancement Classification for Contrast-Enhanced Mammography
Ripaud, Elodie (GE HealthCare), Bloch, Isabelle (GE HealthCare)
ClassificationOptimizationConvolutional Neural NetworkSupervised Fine-TuningImageMagnetic Resonance Imaging
🎯 What it does: A two-stage deep learning framework is proposed, which explicitly models the reader-specific embeddings to assess the background gland enhancement (BPE) differences among readers in CEM images, and achieves rapid adaptation for new readers through a small number of actively learned calibration samples.
Deep Learning-based Alignment Measurement in Knee Radiographs
Hu, Zhisen (University of Manchester), Lindner, Claudia (University of Manchester)
Convolutional Neural NetworkImageMagnetic Resonance Imaging
🎯 What it does: Automated localization of anatomical markers in the knee joint and measurement of the knee joint alignment angle (aTFA) in pre- and post-operative AP knee X-rays.
DeepAf: One-Shot Spatiospectral Auto-Focus Model for Digital Pathology
Yeganeh, Yousef (Technische Universitaet Muenchen), Farshad, Azade (Technische Universitaet Muenchen)
ClassificationOptimizationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: This paper proposes a single-frame adaptive focusing model called DeepAf, which is integrated into an automated microscope system to achieve rapid scanning and automatic focusing of pathological slides.
DEFUSE-MS: Deformation Field-Guided Spatiotemporal Graph-Based Framework for Multiple Sclerosis New Lesion Detection
Salem, Mostafa (Mohamed Bin Zayed University of Artificial Intelligence), Yaqub, Mohammad (Mohamed bin Zayed University of Artificial Intelligence)
SegmentationAnomaly DetectionGraph Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study proposes DEFUSE-MS, a spatiotemporal heterogeneous graph neural network framework guided by deformation fields for the automatic detection of new T2 hyperintense lesions in multiple sclerosis.
Delving into Out-of-Distribution Detection with Medical Vision-Language Models
Ju, Lie (Monash University), Ge, Zongyuan (Melbourne University)
Anomaly DetectionTransformerPrompt EngineeringVision Language ModelMultimodalityBiomedical DataBenchmark
🎯 What it does: This paper systematically evaluates the performance of medical vision-language models (VLM) in full-spectrum OOD detection and proposes a hierarchical prompt method that combines zero-shot and few-shot fine-tuning to enhance OOD detection effectiveness.
DentEval: Fine-tuning-Free Expert-Aligned Assessment in Dental Education via LLM Agents
Deng, Xinyu (University of Sydney), Liu, Daochang (University of Western Australia)
Large Language ModelPrompt EngineeringTextMultimodalityRetrieval-Augmented Generation
🎯 What it does: This paper presents DentEval, a dental education automatic scoring framework based on LLM agents without fine-tuning.
DetectDiffuse: Aggregation- and Attention-driven Universal Lesion Detection with Multi-scale Diffusion Model
Li, Xinyu (Central South University), Yang, Jian (Bejing Institute of Technology)
Object DetectionConvolutional Neural NetworkDiffusion modelImageMagnetic Resonance Imaging
🎯 What it does: A global lesion detection framework called DetectDiffuse based on a multi-scale diffusion model is designed, utilizing noise boxes for prediction and enhancing detection accuracy through two main modules: neighborhood aggregation and 3D stripe attention.
DGHFA: Dynamic Gradient and Hierarchical Feature Alignment for Robust Distillation of Medical VLMs
Xiao, Boyi (University of Science and Technology of China), Zhang, Shaoting (Zhongda Hospital)
Knowledge DistillationAdversarial AttackTransformerVision Language ModelImageBiomedical Data
🎯 What it does: Robust knowledge distillation for medical visual language models is conducted, proposing a framework called DGHFA that incorporates dynamic gradient calibration and hierarchical adversarial feature alignment.
DGM: Disentangled Generative Model for Detecting AD Individualized Pathological Changes via Pseudo-Healthy Synthesis
Li, Zhuangzhuang (Beijing University of Posts and Telecommunications), Liu, Yong (Beijing University of Posts and Telecommunications)
GenerationData SynthesisAnomaly DetectionGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: This paper proposes a separable generative model (DGM) that can transform pathological MRI images of Alzheimer's disease (AD) patients into pseudo-healthy images and generate corresponding residual (lesion) maps to detect and quantify individualized pathological changes.
DGMIR: Dual-Guided Multimodal Medical Image Registration based on Multi-view Augmentation and On-site Modality Removal
Le, Gao (Chongqing University of Posts and Telecommunications), Gao, Xinbo (Chongqing University of Posts and Telecommunications)
Image TranslationSegmentationConvolutional Neural NetworkMultimodalityBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A dual-guided multi-modal medical image registration framework DGMIR is proposed, achieving high-precision registration through modules such as multi-scale heterogeneous feature fusion, perspective feature reorganization, and modality removal.
DHGFormer: Dynamic Hierarchical Graph Transformer for Disorder Brain Disease Diagnosis
Xue, Rundong (Xi'an Jiaotong University), Du, Shaoyi (Xi'an Jiaotong University)
Graph Neural NetworkTransformerGraphBiomedical DataAlzheimer's Disease
🎯 What it does: A dynamic hierarchical graph Transformer framework called DHGFormer is proposed for generating and analyzing functional brain networks, thereby facilitating the diagnosis of brain diseases.
Dia-LLaMA: Towards Large Language Model-driven CT Report Generation
Chen, Zhixuan (Hong Kong University of Science and Technology), Chen, Hao (Harvard Medical School)
GenerationTransformerLarge Language ModelPrompt EngineeringTextComputed Tomography
🎯 What it does: A large language model framework called Dia-LLaMA based on LLaMA2-7B is proposed to generate CT reports guided by diagnostic prompts.
DiDGen: Diffusion-based Dual-task Synthesis for Dermoscopic Data Generation
Shentu, Junjie (Durham University), Al Moubayed, Noura (Durham University)
ClassificationSegmentationGenerationData SynthesisLarge Language ModelSupervised Fine-TuningDiffusion modelImageBiomedical Data
🎯 What it does: The DiDGen framework is proposed, utilizing diffusion models to simultaneously generate skin images and corresponding lesion masks under a single fine-tuning, addressing the issue of insufficient data.
Diff-RRG: Longitudinal Disease-wise Patch Difference as Guidance for LLM-based Radiology Report Generation
Yun, Hannah (Korea University), Suk, Heung-Il (Korea University)
GenerationExplainability and InterpretabilityTransformerLarge Language ModelPrompt EngineeringImageTextBiomedical DataElectronic Health Records
🎯 What it does: Proposes the Diff-RRG framework, which utilizes longitudinal lesion-level difference maps to provide progression guidance for LLM, thereby generating radiology reports with more clinical semantics and interpretability.
DiffAtlas: GenAI-fying Atlas Segmentation via Image-Mask Diffusion
Zhang, Hantao (Swiss Federal Institute of Technology Lausanne), Fua, Pascal (Stony Brook University)
SegmentationGenerationDomain AdaptationDiffusion modelImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes DiffAtlas, which utilizes an image-mask diffusion model to jointly learn the joint distribution of images and segmentation masks during training, achieving Atlas segmentation based on generative AI.
Difficulty Estimation for Image-Specific Medical Image Segmentation Quality Control
Fournel, Joris (DTU Compute), Feragen, Aasa (DTU Compute)
SegmentationConvolutional Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: Proposes a segmentation quality threshold based on image-specific criteria for quality control in medical image segmentation;
DiffOSeg: Omni Medical Image Segmentation via Multi-Expert Collaboration Diffusion Model
Zhang, Han (Sichuan University), Li, Kang (Shanghai Artificial Intelligence Laboratory)
SegmentationPrompt EngineeringDiffusion modelImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes DiffOSeg, a two-stage diffusion model that can simultaneously learn the consensus segmentation results of multiple experts and the individual preference segmentation results.
DiffStain: Conditioned Diffusion-Based Semantic Virtual Staining with Mask Guidance
Han, Yikai (Beihang University), Pei, Yuru (Peking University)
GenerationData SynthesisDiffusion modelImage
🎯 What it does: This study investigates a virtual staining method based on a conditional diffusion model that utilizes deep spectral clustering to generate high-quality multi-channel fluorescence images.
Diffusing Boundaries: CBCT-to-CT Translation with Extended Field of View
Spinat, Quentin (TheraPanacea), Komodakis, Nikos (TheraPanacea)
Image TranslationData SynthesisDiffusion modelImageComputed Tomography
🎯 What it does: A CBCT→CT translation framework based on diffusion models has been developed, utilizing planning CT for boundary expansion and filling to generate synthetic CT of complete scenes.
Diffusion-based Multi-modal MR Fusion for TOF-MRA Image Synthesis
Yu, Tianen, Zhou, Tao (Nanjing University of Science and Technology)
GenerationData SynthesisConvolutional Neural NetworkDiffusion modelImageMultimodalityMagnetic Resonance Imaging
🎯 What it does: A multi-modal diffusion model (MMDM) is proposed to synthesize high-quality TOF-MRA images from multi-modal brain MRI (T1W, T2W, FLAIR).
Diffusion-Based User-Guided Data Augmentation for Coronary Stenosis Detection
Seo, Sumin (Medipixel, Inc.), Jung, Chung-Hwan (Medipixel, Inc.)
ClassificationObject DetectionData SynthesisConvolutional Neural NetworkDiffusion modelImageBiomedical Data
🎯 What it does: This paper proposes a user-guided data augmentation method based on diffusion models (DiGDA), which generates realistic narrow lesions in coronary angiography images by controlling vascular segmentation masks, original images, and text prompts, thereby achieving precise control over lesion severity (%DS). The synthetic images are used alongside real images to train a single-stage YOLO detection and severity classification model.
Diffusion-based Virtual Staining from Polarimetric Mueller Matrix Imaging
Zheng, Xiaoyu (Queen Mary University of London), Chen, Hao (Harvard Medical School)
Image TranslationData SynthesisDiffusion modelAuto EncoderImageMultimodality
🎯 What it does: This paper proposes a virtual staining method based on a bridge diffusion model (RBDM), which maps Mueller matrix polarized images to H&E and fluorescence stained images, and constructs the first multimodal polarized pathology dataset.
DIGS: Dynamic CBCT Reconstruction using Deformation-Informed 4D Gaussian Splatting and a Low-Rank Free-Form Deformation Model
Huang, Yuliang (University College London), McClelland, Jamie R. (University College London)
Gaussian SplattingImageComputed Tomography
🎯 What it does: A 4D Gaussian splatting method based on deformation awareness (DIGS) is proposed for real-time reconstruction of dynamic CBCT images on each projection.
DINO Adapted to X-Ray (DAX): Foundation Models for Intraoperative X-Ray Imaging
Scheuplein, Joshua (Friedrich-Alexander-Universität), Kreher, Björn (Friedrich-Alexander-Universität)
ClassificationObject DetectionSegmentationConvolutional Neural NetworkTransformerContrastive LearningImageBiomedical Data
🎯 What it does: A large-scale intraoperative X-ray image dataset was constructed, and the DINO self-supervised learning framework was adapted as the DAX base model. Subsequently, evaluations were conducted on three downstream tasks: body region classification, metal implant segmentation, and screw detection.
Direct Inversion Formula of the Multi-coil MR Operator under Arbitrary Trajectories
Chen, Junzhou (University of Southern California), Fan, Zhaoyang (University of Southern California)
ImageMagnetic Resonance Imaging
🎯 What it does: This paper proposes a direct inversion formula for multi-coil MRI forward operators for arbitrary sampling trajectories, utilizing low displacement rank (LDR) theory to achieve efficient inversion.
Directional Adaptive Shuffle-Based Visual State-Space Models for Medical Image Restoration
Chan, Simon C. K. (Hong Kong University of Science and Technology), Wong, Terence T. W. (Hong Kong University of Science and Technology)
RestorationSuper ResolutionConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed TomographyPositron Emission Tomography
🎯 What it does: This paper proposes a directionally adaptive shuffling-based Mamba model, DASMamba, specifically designed for medical image restoration. It achieves super-resolution, denoising, and synthesis of low-quality MRI, CT, and PET images through a U-shaped hierarchical structure combined with a state space model.
DISCLOSE the Neurodegeneration Dynamics: Individualized ODE Discovery for Alzheimer’s Disease Precision Medicine
Jung, Wooseok (VUNO Inc.), Kim, Won Hwa (Pohang University of Science and Technology)
Time SeriesBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's DiseaseOrdinary Differential Equation
🎯 What it does: This paper proposes the DISCLOSE framework, which starts from baseline Alzheimer's disease-related information (such as the number of APOE4 alleles and β-amyloid load) and uses an individualized ordinary differential equation (ODE) model to predict the trajectory of brain region atrophy in MCI patients over time.
DisDiff: Disentanglement Diffusion Network for MR Imaging Translation
Zhang, Yipin, Zhang, Xiao-Yong (Fudan University)
Image TranslationGenerationDiffusion modelGenerative Adversarial NetworkImageMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The DisDiff model is proposed for multimodal MR image translation, achieving structural preservation and high-quality generation by decoupling content and attributes.
Disentangled and Interpretable Multimodal Attention Fusion for Cancer Survival Prediction
Eijpe, Aniek (Utrecht University), Silva, Wilson (Utrecht University)
Explainability and InterpretabilityRepresentation LearningTransformerMultimodalityBiomedical Data
🎯 What it does: A multi-modal fusion framework based on attention, DIMAF, is proposed to integrate whole slide images (WSI) and transcriptomic data to predict cancer survival time, achieving representation decoupling through explicit separation of intra-modal and cross-modal interactions.
Distilling foundation models for robust and efficient models in digital pathology
Filiot, Alexandre (Owkin Inc), Olivier, Antoine (Owkin)
CompressionComputational EfficiencyKnowledge DistillationTransformerContrastive LearningImageBiomedical Data
🎯 What it does: Using a teacher-student distillation approach, a digital pathology base model H-Optimus-0 with over one billion parameters was distilled into a ViT-Base model (H0-mini) with only 86M parameters, achieving significant compression in model size and inference cost.
Distribution-Guided Multi-Tracer Brain PET Synthesis from Structural MRI with Class-Conditioned Weighted Diffusion
Yu, Minhui (University of North Carolina at Chapel Hill), Liu, Mingxia (University of North Carolina at Chapel Hill)
GenerationData SynthesisDiffusion modelFlow-based ModelAuto EncoderImageBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease
🎯 What it does: A distribution-guided diffusion framework (NDF) is proposed to synthesize multiple PET images from a single structural MRI, utilizing a class-conditional weighted diffusion model and a pre-trained regularization flow to achieve high-quality, multi-tracer consistent PET generation.
DIY Challenge Blueprint: From Organization to Technical Realization in Biomedical Image Analysis
Klausmann, Leonard (OTH Regensburg), Palm, Christoph (OTH Regensburg)
Object DetectionSegmentationPose EstimationImageBiomedical Data
🎯 What it does: This paper proposes and implements a DIY (self-built) challenge platform blueprint for biomedical image analysis, validated in the PhaKIR end-view challenge at MICCAI 2024.
Does Connectome Harmonic Analysis pass the Spin Test?
Vock, Raphaël (Université Paris-Saclay), Duchesnay, Edouard (Université Paris-Saclay)
Biomedical DataMagnetic Resonance ImagingDiffusion Tensor Imaging
🎯 What it does: This paper evaluates the dependence on anatomical priors of connection-based harmonic analysis (CHA) in reconstructing resting-state fMRI signals by improving the Spin Test method.
Domain Generalization for Mammogram Classification by Suppressing Domain-Specific Features
Chen, Jiqun, Liu, Baodi (Shandong First Medical University)
ClassificationDomain AdaptationConvolutional Neural NetworkImage
🎯 What it does: This paper proposes a method called MC-SDS that enhances the generalization ability of breast X-ray image classification by suppressing domain features, aiming to address the domain shift problem caused by different devices.
Domain Generalization for Pulmonary Nodule Detection via Distributionally-Regularized Mamba
Lan, Tianzhong, Zhu, Min (Sichuan University)
Object DetectionDomain AdaptationConvolutional Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes the Distribution Regularized Mamba Network (DRMNet) to address the domain generalization problem in lung nodule detection.
Domain-Adaptive Diagnosis of Lewy Body Disease with Transferability Aware Transformer
Yu, Xiaowei (University of Texas at Arlington), Zhu, Dajiang (University of Georgia)
ClassificationDomain AdaptationTransformerBiomedical DataAlzheimer's Disease
🎯 What it does: This paper proposes a domain adaptation method based on visual Transformer, TAT, for diagnosis in the context of Lewy Body Disease (LBD) where data is scarce and domain shift exists.
Domain-Agnostic Stroke Lesion Segmentation Using Physics-Constrained Synthetic Data
Chalcroft, Liam (University College London), Ashburner, John (University College London)
SegmentationData SynthesisDomain AdaptationConvolutional Neural NetworkImageMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: By combining quantitative MRI parameters and physical forward models, two physics-constrained synthetic data methods (qATLAS and qSynth) are proposed and used to train stroke lesion segmentation models to enhance cross-domain generalization capabilities.
DpDNet: An Dual-Prompt-Driven Network for Universal PET-CT Segmentation
Liang, Xinglong (Netherlands Cancer Institute), Mann, Ritse (Nanjing University of Information Science and Technology)
SegmentationConvolutional Neural NetworkPrompt EngineeringImageBiomedical DataComputed TomographyPositron Emission Tomography
🎯 What it does: A dual-prompt driven network DpDNet was designed and implemented for the unified segmentation of multiple cancer types in whole-body tumors from PET-CT.
DPGS-Net: Dual Prior-Guided Cross-Domain Adaptive Framework for Ultrasound Image Segmentation
Zhang, Weijie (Guangdong University of Foreign Studies), Gong, Yongyi (Sun Yat-sen University)
SegmentationDomain AdaptationConvolutional Neural NetworkImageUltrasound
🎯 What it does: A dual prior-guided two-stage cross-domain adaptation framework DPGS-Net is proposed for few-shot ultrasound image segmentation.
DSFC: Deformation-Aware Learning Strategy via Self-sustaining Feedback Cycle for Medical Vision Foundation Model Domain Adaptation
Lin, Jie (Xiamen University), Wang, Liansheng (Shanghai Changhai Hospital)
SegmentationDomain AdaptationImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The DSFC framework is proposed, which adapts the basic model of medical imaging to the domain through global and local deformation perturbations and a self-feedback loop, and introduces hard sample adaptive loss to enhance training stability.
Dual Correlation-aware Mamba for Microvascular Obstruction Identification in Non-contrast Cine Cardiac Magnetic Resonance
Yan, Yige (Nanyang Technological University), Rajapakse, Jagath C. (Nanyang Technological University)
RecognitionAnomaly DetectionConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A dual correlation self-attention Mamba model is proposed, utilizing neighboring frame correlation and diastolic frame correlation to identify microvascular occlusion in non-contrast cardiac magnetic resonance imaging.
Dual Knowledge-Aware Guidance for Source-Free Domain Adaptive Fundus Image Segmentation
Chen, Yu (Shanghai Key Laboratory of Trustworthy Computing, East China Normal University), Cao, Guitao (Ant Group)
SegmentationDomain AdaptationImage
🎯 What it does: This paper proposes a Dual Knowledge-Guided (DKG) method for source-independent domain adaptation of retinal image segmentation, enhancing segmentation performance in the target domain.
Dual Selective Gleason Pattern-Aware Multiple Instance Learning for Grade Group Prediction in Histopathology Images
Hao, Xinyu (Dalian University of Technology), Cong, Fengyu (University of Jyvaskyla)
ClassificationKnowledge DistillationTransformerImageBiomedical Data
🎯 What it does: A dual-selection aggregation multi-instance learning framework, DSPA-MIL, is proposed, which uses learnable aggregation symbols and expert concept-guided aggregation to jointly achieve patient-level Grade Group prediction.
Dual-Branch Dynamic Coupling Weakly Supervised Learning for Class-Incremental Histopathological Region Segmentation
Hong, Xiaoyan (Jiangnan University), Pan, Xiang (Hangzhou Dianzi University)
SegmentationKnowledge DistillationTransformerImage
🎯 What it does: This paper proposes the DDCWISS network, which utilizes weakly supervised image-level labels to achieve class-incremental segmentation of pathological regions, and reduces the cost of pixel-level annotations and catastrophic forgetting through dual-branch dynamic coupling and dual-path supervision.
Dual-Stream Multi-Band Fusion Network for Dynamic Functional Connectivity Analysis in Brain Disorder Classification
Wu, Ling (Beijing Normal University), Guo, Xiaojuan (Beijing Normal University)
ClassificationRecurrent Neural NetworkGraph Neural NetworkTime SeriesBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A Dual-Stream Multi-Frequency Fusion Network (DSMFN) is proposed for dynamic functional connectivity analysis to achieve classification of brain disorders (MCI, ASD).
DualPrompt-MedCap: A Dual-Prompt Enhanced Approach for Medical Image Captioning
Zhao, Yining (University of Technology Sydney), Braytee, Ali (University of Technology Sydney)
RecognitionGenerationConvolutional Neural NetworkSupervised Fine-TuningPrompt EngineeringVision Language ModelImageTextMultimodalityMagnetic Resonance Imaging
🎯 What it does: This paper proposes DualPrompt-MedCap, a dual-prompt medical image captioning framework that combines modality awareness and question guidance.
DuoDent: Tooth Generation using Dual-Stream Diffusion with Normal Consistency
Kwon, Doeyoung (Korea University), Baek, Seung Jun (Korea University Anam Hospital)
GenerationData SynthesisConvolutional Neural NetworkTransformerDiffusion modelPoint CloudMeshComputed Tomography
🎯 What it does: The DuoDent framework is proposed, utilizing a dual-stream diffusion model to generate high-precision dental point clouds, and reconstructing smooth dental meshes through regularized normal consistency and directional consistency.
DyMAS-Net: Dynamic Multi-Scale Adaptive Sampling Network for Efficient Medical Image Segmentation
Wang, Siqi (Nankai University), Tian, Jun (Nankai University)
SegmentationConvolutional Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: A lightweight medical image segmentation network DyMAS-Net is designed, combining Hierarchical Multi-Scale Convolution Blocks (HMCB), Adaptive Dynamic Sampling Module (ADSM), and Dual Attention Fusion Unit (DAFU) to achieve high-precision segmentation.
Dyna3DGR: 4D Cardiac Motion Tracking with Dynamic 3D Gaussian Representation
Fu, Xueming (University of Science and Technology of China), Zhou, S. Kevin (Hohai University)
Object TrackingSegmentationGaussian SplattingBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A self-supervised optimization framework Dyna3DGR is proposed, which combines explicit 3D Gaussian representation with implicit neural motion fields to estimate 4D cardiac motion in real-time;
Dynamic Function-Structure Connectivity Coupling for Predicting Progression Trajectories in Neurocognitive Decline
Wang, Qianqian (University of North Carolina at Chapel Hill), Liu, Mingxia (University of North Carolina at Chapel Hill)
Graph Neural NetworkTransformerBiomedical DataMagnetic Resonance ImagingDiffusion Tensor ImagingAlzheimer's Disease
🎯 What it does: A dynamic functional-structural coupling framework (DFSC) based on fMRI and DTI is proposed to predict the progression trajectory of neurocognitive decline.
Dynamic Gradient Sparsification Training for Few-Shot Fine-tuning of CT Lymph Node Segmentation Foundation Model
Luo, Zihao (University of Electronic Science and Technology of China), Luo, Xiangde (Stanford University)
SegmentationConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a LN segmentation baseline model using Dynamic Gradient Sparsification Training (DGST) under few-shot conditions, enhancing the segmentation performance of CT neck and mediastinal LN through pre-training and fine-tuning.
Dynamic-Aware Spatio-temporal Representation Learning for Dynamic MRI Reconstruction
Baik, Dayoung (Ulsan National Institute of Science and Technology), Yoo, Jaejun (Ulsan National Institute of Science and Technology)
RestorationRepresentation LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A Dynamic Aware Implicit Neural Representation (DA-INR) framework is proposed for unsupervised dynamic MRI reconstruction, achieving fast convergence and high-quality reconstruction by combining hash coding and normalized space modeling.
E-BayesSAM: Efficient Bayesian Adaptation of SAM with Self-Optimizing KAN-Based Interpretation for Uncertainty-Aware Ultrasonic Segmentation
Huang, Bin (Shenzhen University), Li, Shuo (Harbin Institute of Technology)
SegmentationComputational EfficiencyImageBiomedical DataUltrasound
🎯 What it does: An efficient Bayesian Segment Anything Model (E-BayesSAM) is proposed for unsupervised ultrasound image segmentation and uncertainty estimation.
EchoCardMAE: Video Masked Auto-Encoders Customized for Echocardiography
Yang, Xuan (Shenzhen University), Chen, Yen-Wei (Ritsumeikan University)
SegmentationRepresentation LearningTransformerAuto EncoderContrastive LearningVideoBiomedical DataUltrasound
🎯 What it does: This paper proposes EchoCardMAE, a self-supervised masked video autoencoder specifically designed for cardiac ultrasound videos, aimed at learning general features from unlabeled data.
EchoingECG: An Electrocardiogram Cross-Modal Model for Echocardiogram Tasks
Gao, Yuan (University Health Network), McIntosh, Chris (University Health Network)
Contrastive LearningTime SeriesBiomedical DataUltrasoundElectrocardiogram
🎯 What it does: A probabilistic student-teacher model called EchoingECG has been developed to predict echocardiographic (ECHO) related metrics using ECG signals.
EchoViewCLIP: Advancing Video Quality Control through High-performance View Recognition of Echocardiography
Song, Shanshan (Hong Kong University of Science and Technology), Li, Xiaomeng (Southern Medical University)
RecognitionAnomaly DetectionVision Language ModelContrastive LearningVideoMultimodalityUltrasound
🎯 What it does: This paper proposes the EchoViewCLIP framework, achieving fine recognition of 38 categories of cardiac ultrasound views, and integrates OOD detection and quality assessment.
Edge-Aware Hierarchical Graph Transformer to Decode Brain Arterial Network
Zhang, Kaiyu (University of Washington), Yuan, Chun (University of Washington)
ClassificationAnomaly DetectionGraph Neural NetworkTransformerMultimodalityGraphBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A hierarchical graph transformer (HGT) aimed at the brain arterial network (BAN) was developed, and a multimodal dataset was constructed that includes a whole-brain BAN graph, vascular geometric features, and clinical indicators for predicting various cardiovascular risk factors.
Edge-Aware Token Halting for Efficient and Accurate Medical Image Segmentation
Guo, Yuhao (Institute of Intelligent Machines, Chinese Academy of Sciences), Cheng, Erkang (Institute of Intelligent Machines, Chinese Academy of Sciences)
SegmentationTransformerImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes an efficient segmentation framework HRViT based on Vision Transformer, which achieves high-precision semantic segmentation of medical images through dynamic edge token stopping and reconstruction.
Edge-semantic Synergy Fusion and Adaptive Noise-aware for Weakly Supervised Pathological Tissue Segmentation
Zhang, Hualong (Guilin University of Electronic Technology), Pan, Xipeng (Guilin University of Electronic Technology)
SegmentationTransformerImageBiomedical Data
🎯 What it does: A weakly supervised pathological tissue segmentation framework called ESFAN is proposed, which generates high-quality pseudo-masks through edge semantic collaborative fusion and enhances segmentation performance using an adaptive noise-aware mechanism.
EdgeANet: A Transformer-based Edge Representation Learning Network for Canine X-ray Verification
Lee, In-Gyu (Chungbuk National University), Jeong, Ji-Hoon (Chungbuk National University)
Object DetectionRepresentation LearningTransformerImage
🎯 What it does: This paper proposes EdgeANet, a Transformer-based edge representation learning network for the validation of rotated vertebrae in canine chest X-ray images.
EEG-DINO: Learning EEG Foundation Models via Hierarchical Self-Distillation
Wang, Xujia (Central Research Institue, United Imaging Healthcare, Co., Ltd.), Zhen, Xiantong (Central Research Institue, United Imaging Healthcare, Co., Ltd.)
Knowledge DistillationRepresentation LearningTransformerTime SeriesBiomedical Data
🎯 What it does: Proposes EEG-DINO, a hierarchical self-distillation based EEG foundational model that learns multi-view semantic alignment and spatial-temporal decoupled encoding.
EFFDNet: A Scribble-Supervised Medical Image Segmentation Method with Enhanced Foreground Feature Discrimination
Liu, Jinhua (Nanyang Technological University), Yeo, Si Yong (Institute for Infocomm Research)
SegmentationConvolutional Neural NetworkContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a handwriting-supervised medical image segmentation method called EFFDNet, which enhances foreground recognition capability by utilizing the foreground-background semantics in handwriting.
EfficientMedNeXt: Multi-Receptive Dilated Convolutions for Medical Image Segmentation
Rahman, Md Mostafijur (University of Texas at Austin), Marculescu, Radu (University of Texas at Austin)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A lightweight, end-to-end 3D medical image segmentation architecture called EfficientMedNeXt is proposed, which optimizes the encoder-decoder structure of MedNeXt in two stages, removing high-resolution redundancy and unifying decoder channels.
EFMS-Net: Efficient Frequency-Enhanced Multi-Scale Network for Ischemic Stroke Segmentation
Yang, Jie, Zhan, Yihong (First Affiliated Hospital of Xiamen University)
SegmentationConvolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An efficient frequency domain enhanced multi-scale network (EFMS-Net) is proposed for 3D segmentation of ischemic stroke infarct areas in the brain.
EG-Net: An Edge-Guided Network for Rigid Registration of Laparoscopic Low-Overlap Point Clouds
Wu, Wenbin (University of Science and Technology of China), Gao, Xin (Chinese Academy of Sciences)
OptimizationTransformerPoint Cloud
🎯 What it does: A marginally guided network EG-Net is proposed for rigid registration of low-overlap laparoscopic point clouds.
Eliminating Language Bias for Medical Visual Question Answering with Counterfactual Contrastive Training
Wan, Xingyu, Liu, Zhe (Shenzhen University)
ClassificationRecognitionTransformerContrastive LearningImageMultimodalityBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: The DeCoCT method is proposed, which decomposes the causal relationship between clinical terms and answers in medical visual question answering (Med-VQA), uses adversarial contrastive training to eliminate language bias, and introduces a Key Region Capture Module (KRCM) to enhance the model's focus on visual key information.
Embedding-Based Federated Data Sharing via Differentially Private Conditional VAEs
Di Salvo, Francesco (University of Bamberg), Ledig, Christian (University of Bamberg)
GenerationData SynthesisFederated LearningAuto EncoderGenerative Adversarial NetworkImageBiomedical DataComputed Tomography
🎯 What it does: A federated data sharing method based on Differential Privacy Conditional Variational Autoencoder (DP‑CVAE) is proposed, which trains a generator on the feature embeddings extracted from the base model to generate synthetic data usable for multiple tasks;
End-to-End 3D Tooth Landmark Detection with Fuzzy Tooth Localization
Shi, Kaibo (Zhejiang University), Zheng, Youyi (Zhejiang University)
Object DetectionGraph Neural NetworkTransformerPoint CloudMesh
🎯 What it does: An end-to-end two-stage 3D tooth landmark detection framework, ToothLDNet, is proposed, which can directly locate teeth and detect landmarks from 3D mandibular models without the need for tooth segmentation labels.
Endo-4DGX: Robust Endoscopic Scene Reconstruction and Illumination Correction with Gaussian Splatting
Huang, Yiming (Chinese University of Hong Kong), Ren, Hongliang (Chinese University of Hong Kong)
RestorationDepth EstimationGaussian SplattingImage
🎯 What it does: In the dynamic endoscopic scene, we propose Endo-4DGX, which combines illumination embedding, region-aware enhancement, and spatial-aware correction within a Gaussian expansion framework to achieve 3D reconstruction and illumination correction under uneven lighting.
Endo-CLIP: Progressive Self-Supervised Pre-training on Raw Colonoscopy Records
He, Yili (Fudan University), Wang, Shuo (Fudan University)
ClassificationObject DetectionSegmentationTransformerLarge Language ModelContrastive LearningImageTextMultimodalityElectronic Health Records
🎯 What it does: This paper proposes Endo-CLIP, a progressive self-supervised pre-training framework for aligning images and text based on raw colonoscopy recordings.
Endo-FASt3r: Endoscopic Foundation model Adaptation for Structure from motion
Sheikh Zeinoddin, Mona (University College London), Stoyanov, Danail (University College London)
Pose EstimationDepth EstimationDomain AdaptationTransformerContrastive LearningImage
🎯 What it does: Developed Endo-FASt3r, a self-supervised learning-based monocular depth and pose estimation framework, utilizing two foundational models (Depth Anything V2 and Reloc3rX) for end-to-end estimation.
Endo-GSMT: Endoscopic Monocular Scene Reconstruction with Dynamic Gaussian Splatting and Motion Tracking
Gou, Hao (Hangzhou Dianzi University), Luo, Huoling (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences)
Gaussian SplattingSimultaneous Localization and MappingVideo
🎯 What it does: A dynamic surgical scene reconstruction method based on monocular endoscopic video is proposed, utilizing 3D Gaussian Splatting to achieve high-quality 3D reconstruction.
Endo3R: Unified Online Reconstruction from Dynamic Monocular Endoscopic Video
Guo, Jiaxin (Chinese University of Hong Kong), Liu, Yun-Hui (Chinese University of Hong Kong)
Pose EstimationDepth EstimationTransformerOptical FlowVideoPoint Cloud
🎯 What it does: We propose Endo3R, a unified online 3D reconstruction framework based on monocular endoscopic video, capable of simultaneously outputting global point clouds, scale-consistent depth maps, as well as camera poses and intrinsic parameters.
EndoDAV: Depth Any Video in Endoscopy with Spatiotemporal Accuracy
Zhou, Zanwei (Shanghai Jiao Tong University), Shen, Wei (Zhejiang University)
Depth EstimationSupervised Fine-TuningVideo
🎯 What it does: Perform parameter-efficient fine-tuning of Video Depth Anything on endoscopic videos to achieve spatial accuracy and temporal consistency in depth estimation.
EndoFlow-SLAM: Real-Time Endoscopic SLAM with Flow-Constrained Gaussian Splatting
Wu, Taoyu (Xi'an Jiaotong Liverpool University), Li, Haoang (Hong Kong University of Science and Technology (Guangzhou))
Pose EstimationDepth EstimationOptimizationGaussian SplattingSimultaneous Localization and MappingOptical FlowBiomedical Data
🎯 What it does: A real-time endoscopic SLAM system (EndoFlow-SLAM) based on 3D Gaussian Splatting with optical flow constraints is proposed, achieving accurate camera pose estimation and high-quality dense reconstruction.
EndoGen: Conditional Autoregressive Endoscopic Video Generation
Liu, Xinyu (Chinese University of Hong Kong), Yuan, Yixuan (Southern University of Science and Technology)
GenerationData SynthesisGenerative Adversarial NetworkVideo
🎯 What it does: EndoGen is proposed, a conditional autoregressive endoscopic video generation framework that can generate high-quality, temporally consistent endoscopic videos based on pathological categories.
EndoMamba: An Efficient Foundation Model for Endoscopic Videos via Hierarchical Pre-training
Tian, Qingyao (Chinese Academy of Sciences), Liu, Hongbin (Chinese Academy of Sciences)
ClassificationRecognitionSegmentationKnowledge DistillationTransformerAuto EncoderVideo
🎯 What it does: This paper presents EndoMamba, a foundational model based on the Mamba state space model specifically designed for real-time endoscopic video analysis, and enhances performance in end-to-end tasks through hierarchical self-supervised pre-training.
EndoMetric: Near-Light Monocular Metric Scale Estimation in Endoscopy
Iranzo, Raúl (Universidad de Zaragoza), Montiel, José M. M. (Universidad de Zaragoza)
Depth EstimationOptimizationSimultaneous Localization and MappingImageVideoMagnetic Resonance Imaging
🎯 What it does: Estimate the true scale using a near-light illumination model on monocular endoscopic images, completing 3D reconstruction and measurement from scale-free to true scale.
EndoPlanar: Deformable Planar-based Gaussian Splatting for Surgical Scene Reconstruction
Paonim, Thatphum (Chulalongkorn University), Vateekul, Peerapon (Chulalongkorn University)
RestorationSegmentationDepth EstimationGaussian SplattingVideoBenchmark
🎯 What it does: A deformable planar-based Gaussian scattering method called EndoPlanar is proposed for real-time reconstruction of soft tissues in endoscopic stereo videos.
Endoscopic Artifact Inpainting for Improved Endoscopic Image Segmentation
Yu, Zhangyuan (Beijing University of Posts and Telecommunications), Lao, Qicheng (Ningbo Fregty Optoelectronics Technology Co., Ltd)
RestorationSegmentationConvolutional Neural NetworkImage
🎯 What it does: A two-stage framework for removing artifacts from endoscopic images is proposed, which first suppresses specular reflections and then fills in the diffuse reflection areas to enhance segmentation performance.
Endoscopic Depth-of-Field Expansion via Cascaded Network with Two-streamed Multi-scale Fusion
Deng, Xiang (Zhejiang University), Ye, Xuesong (Zhejiang University)
Image TranslationRestorationConvolutional Neural NetworkTransformerImage
🎯 What it does: A dual-stream cascade encoding-decoding network is proposed to achieve multi-focus image fusion for endoscopy, thereby extending the depth of field.
Endplate3D-QCT: A High-Resolution Dataset and Benchmark for Automated 3D Segmentation of Lumbar Vertebral Endplates in QCT
Yin, Zixun (Peking University), Wang, Ping (Peking University)
SegmentationConvolutional Neural NetworkImageBiomedical DataComputed TomographyBenchmark
🎯 What it does: The first publicly available high-resolution lumbar endplate 3D QCT dataset, Endplate3D-QCT, has been proposed, along with pixel-level annotations and an evaluation framework.
Enforcing Geometric Constraints of Surface Normal and Pose for Self-supervised Monocular Depth Estimation on Laparoscopic Images
Li, Wenda (Nagoya University), Mori, Kensaku (Aichi Institute of Technology)
Pose EstimationDepth EstimationVideo
🎯 What it does: This paper proposes a self-supervised monocular depth estimation method that enhances depth estimation quality on laparoscopic images by introducing surface normal estimation, distance uncertainty, and a 4D score volume to achieve geometric constraints.
Enhancing AI-assisted Stroke Emergency Triage with Adaptive Uncertainty Estimation
Yang, Shuhua (Pennsylvania State University), Wong, Stephen T. C. (Pennsylvania State University)
ClassificationAnomaly DetectionConvolutional Neural NetworkSupervised Fine-TuningMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An adaptive uncertainty estimation network based on video and audio multimodal data, AUSTIN, is proposed to enhance the accuracy and reliability of stroke triage in the emergency room.