MICCAI 2025 Papers — Page 9
International Conference on Medical Image Computing and Computer-Assisted Intervention · 1027 papers
SAMASK-CLTR: A spatial-aware mask guided learning model for benign and malignant tumor classification in ABUS
Xu, Peirong (Macao Polytechnic University), Tan, Tao (Macao Polytechnic University)
ClassificationConvolutional Neural NetworkTransformerImageUltrasound
🎯 What it does: A hybrid network SAMASK-CLTR based on ResNet-50 and Transformer is proposed, achieving binary classification of benign and malignant tumors in ABUS images through mask prompts with 3D spatial position encoding.
SAMed-2: Selective Memory Enhanced Medical Segment Anything Model
Yan, Zhiling (Lehigh University), Sun, Lichao (Lehigh University)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper presents SAMed-2, a segmentation foundation model based on SAM-2 for medical imaging;
SAMSA: Segment Anything Model enhanced with Spectral Angles for Hyperspectral Interactive Medical Image Segmentation
Roddan, Alfie (Imperial College London), Giannarou, Stamatia (Imperial College London)
SegmentationImageBiomedical Data
🎯 What it does: Proposes the SAMSA framework, which combines the Segment Anything model with spectral angle similarity to achieve interactive segmentation of hyperspectral medical images.
SAMUSA: Segment Anything Model 2 for UltraSound Annotation
Podvin, Baptiste (Ircad France), Hostettler, Alexandre (Ircad France)
SegmentationImageVideoBiomedical DataUltrasound
🎯 What it does: This paper presents SAMUSA, an interactive ultrasound image/video segmentation tool based on SAM2, which improves segmentation using boundary point and temporal point prompts.
Scalp Diagnostic System With Label-Free Segmentation and Training-Free Image Translation
Kim, Youngmin (Yonsei University), Noh, Junhyug (Ewha Womans University)
Image TranslationSegmentationGenerationConvolutional Neural NetworkDiffusion modelImage
🎯 What it does: This paper presents ScalpVision, a complete diagnostic system for scalp diseases that combines unlabeled hair segmentation and generative data augmentation.
Seeing Beyond the Surface: Retinal Thickness Prediction from Color Fundus Photography for DME Management
Cheng, Wenquan (Tsinghua University), Song, Su Jeong (Sungkyunkwan University)
ClassificationSegmentationConvolutional Neural NetworkTransformerDiffusion modelImage
🎯 What it does: Using color fundus photography to predict retinal thickness maps, assisting in the management of diabetic macular edema (DME)
Segmenting Vessels Encapsulating Tumor Clusters via Fine-Grained Visual Prompt
Yu, Jiahui (Zhejiang University), Xu, Yingke (University of British Columbia)
SegmentationTransformerContrastive LearningImageBiomedical Data
🎯 What it does: This paper proposes a VPP2P framework based on pixel-level visual prompts, utilizing nuclear location priors to achieve semi-supervised fine-grained segmentation of VETC.
Segmentor-guided Counterfactual Fine-Tuning for Locally Coherent and Targeted Image Synthesis
Xia, Tian (Imperial College London), Glocker, Ben (Imperial College London)
SegmentationGenerationData SynthesisSupervised Fine-TuningGenerative Adversarial NetworkImageBiomedical DataComputed Tomography
🎯 What it does: A structured causal model using a pre-trained segmenter for adversarial fine-tuning (Seg-CFT) is proposed, achieving local and controllable interventions on structure-specific attributes in medical images.
Selective Alignment Transfer for Domain Adaptation in Skin Lesion Analysis
Sultana, Nurjahan (Manchester Metropolitan University), Yap, Moi Hoon (Manchester Metropolitan University)
Domain AdaptationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A fully supervised domain adaptation framework, SAT-DA, is proposed to achieve cross-domain transfer of skin lesion images from microscopy to clinical settings through dynamic feature weighting.
Self is the Best Learner: CT-free Ultra-Low-Dose PET Organ Segmentation via Collaborating Denoising and Segmentation Learning
Ye, Zanting (Southern Medical University), Lu, Lijun (Southern Medical University)
SegmentationConvolutional Neural NetworkBiomedical DataComputed TomographyPositron Emission Tomography
🎯 What it does: This paper proposes an end-to-end learning framework called LDOS for multi-organ segmentation on ultra-low dose PET (LDPET) without the need for CT assistance.
Self-adaptive Vision-Language Model for 3D Segmentation of Pulmonary Artery and Vein
Guo, Xiaotong (National Cancer Center, Chinese Academy Of Medical Sciences And Peking Union Medical College), Meng, Yanda (University Of Exeter)
SegmentationConvolutional Neural NetworkVision Language ModelImageBiomedical DataComputed Tomography
🎯 What it does: The LA-CAF framework is proposed, utilizing pre-trained CLIP for 3D segmentation of pulmonary arteries and veins.
Self-Propagative Multi-Task Learning for Predicting Cardiometabolic Risk Factors
Ko, Seonghyeon (Sungkyunkwan University), Choo, Hyunseung (Sungkyunkwan University)
ClassificationRecognitionOptimizationConvolutional Neural NetworkMixture of ExpertsImageBiomedical Data
🎯 What it does: Using a self-propagating multi-task learning model to simultaneously predict 15 cardiovascular metabolic risk factors from fundus images.
Self-supervised Axial Super-Resolution for Volume Microscopy via Diffusion-Guided Structure Distillation
Chen, Bohao (University of Chinese Academy of Sciences), Chen, Xi (Chinese Academy of Sciences)
RestorationSegmentationSuper ResolutionKnowledge DistillationConvolutional Neural NetworkDiffusion modelBiomedical DataStochastic Differential Equation
🎯 What it does: A self-supervised volumetric super-resolution framework D2R is proposed, which first extracts biological structure priors using a 2D diffusion model, then generates pseudo high-resolution volumes, and finally learns stable structural transformations using a 3D VSR network to achieve axial resolution enhancement.
Self-Supervised Distillation of Legacy Rule-Based Methods for Enhanced EEG-Based Decision-Making
Zhang, Yipeng (University of California Los Angeles), Roychowdhury, Vwani (University of California Los Angeles)
ClassificationKnowledge DistillationRepresentation LearningAuto EncoderTime SeriesBiomedical DataElectrocardiogram
🎯 What it does: A large number of candidate events are generated using a legacy rule-based HFO detector, followed by learning event morphological representations through a self-supervised variational autoencoder (VAE), and weak labels are extracted using latent space clustering. Finally, a classifier is trained to accurately screen candidate events for pathological HFOs.
Self-Supervised Multiview Xray Matching
Dabboussi, Mohamad (Telecom Paris), Gori, Pietro (Milvue)
ClassificationObject DetectionTransformerImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes an unsupervised multi-view X-ray matching and fracture detection framework, which generates DRR from unlabeled CT and automatically constructs corresponding matrices to achieve multi-to-multi matching of multi-view X-rays, using the correspondence information as pre-training and attention guidance to enhance multi-view fracture classification performance.
Self-supervised Normality Learning and Divergence Vector-guided Model Merging for Zero-shot Congenital Heart Disease Detection in Fetal Ultrasound Videos
Saha, Pramit (University of Oxford), Noble, J. Alison (University of Oxford)
Domain AdaptationAnomaly DetectionKnowledge DistillationTransformerVideoBiomedical DataUltrasound
🎯 What it does: A zero-shot congenital heart disease detection framework based on self-supervised video normality learning and divergence vector-guided model merging has been developed.
Semantic Interpolative Diffusion Model: Bridging the Interpolation to Masks and Colonoscopy Image Synthesis for Robust Generalization
Heo, Chanyeong (Myongji University), Jung, Jaehee (Myongji University)
SegmentationGenerationData SynthesisDiffusion modelImage
🎯 What it does: A Semantic Interpolation Diffusion Model (SIDM) is proposed, which generates diverse pairs of endoscopic images and masks by interpolating between lesion masks and background semantic labels to enhance the generalization ability of segmentation models.
Semantic Scene Graph for Ultrasound Image Explanation and Scanning Guidance
Li, Xuesong (TU Munich), Jiang, Zhongliang (Computer Aided Medical Procedures (CAMP), TU Munich)
Explainability and InterpretabilityTransformerLarge Language ModelImageUltrasound
🎯 What it does: This paper proposes the use of semantic scene graphs (SG) and large language models (LLM) for providing interpretability and scanning guidance for ultrasound images.
Semantic-Aware Chest X-ray Report Generation with Domain-Specific Lexicon and Diversity-Controlled Retrieval
Zhang, Baochang (Technical University of Munich), Navab, Nassir (Technische Universität München)
GenerationRetrievalConvolutional Neural NetworkTransformerImageTextMultimodalityComputed TomographyRetrieval-Augmented Generation
🎯 What it does: The DrLS framework is proposed, which generates more accurate chest X-ray reports that align with medical semantics through multimodal retrieval, vocabulary-weighted loss, and sentence semantic loss.
Semantically Consistent Discrete Diffusion for 3D Biological Graph Modeling
Prabhakar, Chinmay (University of Zurich), Menze, Bjoern (University of Zurich)
GenerationData SynthesisGraph Neural NetworkTransformerDiffusion modelGraphBiomedical Data
🎯 What it does: A semantic-consistent discrete diffusion model is designed to generate 3D biological graph networks that satisfy anatomical structure and edge label constraints.
Semi-Supervised Deformation-Free Image-to-Image Translation for Realistic CT Synthesis from CBCT
Han, Ji Yong (Seoul National University), Yi, Won-Jin (Seoul National University)
Image TranslationData SynthesisConvolutional Neural NetworkGenerative Adversarial NetworkContrastive LearningImageMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A two-stage semi-supervised unpaired image translation framework is proposed, which generates high-quality, HU value-accurate CT images from CBCT while maintaining complete anatomical structure and high resolution.
Semi-Supervised Multi-Modal Medical Image Segmentation for Complex Situations
Meng, Dongdong (Peking University), Yan, Xueqing (Peking University Cancer Hospital)
SegmentationConvolutional Neural NetworkContrastive LearningImageMultimodalityMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes a semi-supervised multimodal medical image segmentation framework that enhances segmentation accuracy using limited labeled data through multi-stage feature fusion, modality-aware enhancement, and contrastive mutual learning.
SemiVT-Surge: Semi-Supervised Video Transformer for Surgical Phase Recognition
Li, Yiping (Eindhoven University of Technology), Al Khalil, Yasmina (University Medical Center Utrecht)
RecognitionTransformerContrastive LearningVideo
🎯 What it does: A semi-supervised surgical phase recognition framework based on video Transformer is proposed, integrating temporal consistency regularization and category prototype-based contrastive learning.
Separable tissue representations for attributable risk prediction
Wåhlstrand, Victor (Chalmers University of Technology), Häggström, Ida (Chalmers University of Technology)
SegmentationExplainability and InterpretabilityRepresentation LearningTransformerAuto EncoderImageBiomedical DataComputed Tomography
🎯 What it does: Developed the STRAP framework, which utilizes variable-sized ROI masks to learn the separable representations of tissues such as bone, muscle, and fat in high-resolution CT images for fracture risk prediction and interpretability.
Sequence-Independent Continual Test-Time Adaptation with Mixture of Incremental Experts for Cross-Domain Segmentation
Xu, Dunyuan (Chinese University of Hong Kong), Heng, Pheng-Ann (Chinese University of Hong Kong)
SegmentationDomain AdaptationConvolutional Neural NetworkMixture of ExpertsImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An adaptive framework for continuous testing based on Mixture of Incremental Experts (MoIE) is proposed for cross-domain medical image segmentation, which can maintain stable performance in the case of randomly arriving samples.
SFPFR: Self-supervised Facial Paralysis Face Reconstruction Under Few Views
Qiu, Yaru (Qingdao University of Science and Technology), Sun, Yuanyuan (Qingdao University of Science and Technology)
RestorationGenerationImageVideo
🎯 What it does: A self-supervised framework SFPFR is proposed to reconstruct 3D facial models of patients with facial paralysis using 1-3 views.
SG2VID: Scene Graphs Enable Fine-Grained Control for Video Synthesis
Sivakumar, Ssharvien Kumar (Technical University Darmstadt), Mukhopadhyay, Anirban (Carl Zeiss AG)
RecognitionGenerationData SynthesisGraph Neural NetworkDiffusion modelOptical FlowVideoGraph
🎯 What it does: A scene graph-based diffusion model SG2VID is proposed, achieving fine-grained synthesis of surgical videos and human-machine interaction control.
ShareLink: Neuro-Inspired EEG-based Cross-Subject Emotion Recognition via Shared Bi-hemisphere
Huang, Zixuan (Chinese Academy of Sciences), Miao, Fen (University of Electronic Science and Technology of China)
RecognitionGraph Neural NetworkMixture of ExpertsTime SeriesBiomedical Data
🎯 What it does: The ShareLink model is proposed to achieve cross-subject EEG emotion recognition by enhancing recognition performance through the alignment and interaction of features from the left and right hemispheres.
Shuffle-Diversity Collaborative Federated Learning for Imbalanced Medical Image Analysis
Gao, Wenpeng (South China Agricultural University), Fan, Xiaomao (Shenzhen Technology University)
ClassificationFederated LearningConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: Proposes the FedSDC joint network, which addresses sample imbalance and heterogeneity in medical image classification through shared Body and multi-head Head in federated learning, using Shuffle-Diversity collaboration.
Sim-to-Real Transformer-Based Shape Reconstruction for Automated Orthopedic Fracture Reduction Planning
Yibulayimu, Sutuke (Beihang University), Wang, Yu (Capital Medical University)
TransformerBiomedical DataComputed Tomography
🎯 What it does: Proposes an automated fracture reduction planning framework based on Transformer for fracture shape recovery and recursive registration.
SimCroP: Radiograph Representation Learning with Similarity-driven Cross-granularity Pre-training
Wang, Rongsheng (University of Science and Technology of China), Zhou, S. Kevin (Hohai University)
ClassificationSegmentationRepresentation LearningTransformerAuto EncoderContrastive LearningImageTextMultimodalityBiomedical DataComputed Tomography
🎯 What it does: Through the SimCroP framework, which combines cross-granularity pre-training with similarity-driven alignment, joint pre-training of chest CT and corresponding reports is conducted, significantly improving classification and segmentation performance under multi-task settings.
SIMPLE: Simultaneous Multi-plane Self-supervised Learning for Isotropic MRI Restoration from Anisotropic Data
Benisty, Rotem (Technion Israel Institute of Technology), Freiman, Moti (Technion Israel Institute of Technology)
RestorationConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A self-supervised multi-plane learning framework named SIMPLE has been developed to reconstruct clinically acquired anisotropic MRI into isotropic volumes, enhancing the fidelity of three-dimensional structures.
Single Image Test-Time Adaptation via Multi-View Co-Training
Joshi, Smriti (Universitat de Barcelona), Lekadir, Karim (Institució Catalana de Recerca i Estudis Avançats)
SegmentationDomain AdaptationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: For the test-time domain adaptation of single medical images, a multi-view collaborative training method guided by uncertainty is proposed, allowing the model to perform adaptive learning on a single target image only during inference, thus achieving 3D breast MRI tumor segmentation.
Single-spoke Motion-compensated Dynamic 3D MRI Reconstruction via Neural Representation
Chen, Lixuan (University of Michigan), Park, Jeong Joon (University of Michigan)
RestorationNeural Radiance FieldBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A dynamic 3D MRI reconstruction method called SPINER is proposed, which utilizes single pivot motion compensation.
SlimFormer-3D: A Layer-Adaptive Lightweight Transformer for Efficient 3D Medical Image Segmentation
Hong, Yang, Mo, Jianqing (Guangdong University Of Technology)
SegmentationComputational EfficiencyTransformerBiomedical DataComputed Tomography
🎯 What it does: This paper proposes SlimFormer-3D, a lightweight 3D medical image segmentation model that adaptively prunes the internal TokenMixer and MLP of the Transformer at a hierarchical level.
Small Lesions-aware Bidirectional Multimodal Multiscale Fusion Network for Lung Disease Classification
Yu, Jianxun (Xidian University), Wang, Changmiao (Shenzhen Research Institute of Big Data)
ClassificationConvolutional Neural NetworkMultimodalityTabularBiomedical DataComputed TomographyPositron Emission Tomography
🎯 What it does: A multi-modal multi-scale fusion network (MMCAF-Net) aimed at classifying lung diseases is proposed, which simultaneously processes 3D medical images and clinical tabular data, focusing on capturing small lesion information.
Smarter Self-Distillation: Optimizing the Teacher for Surgical Video Applications
Yamlahi, Amine (German Cancer Research Center), Maier-Hein, Lena (Dresden University of Technology)
RecognitionKnowledge DistillationTransformerVideo
🎯 What it does: A multi-teacher self-distillation framework is proposed for surgical video analysis, which selects multiple rounds of teacher models with different random seeds within the 'practical zone' during the training process and averages their soft labels to enhance the performance of the student model; subsequently, a temporal decoder is added to extract spatiotemporal information.
SMF-Net: Unlocking Multimodal Insights for Enhanced Stroke Lesion Segmentation
Atlaw, Meklit (Northwestern Polytechnical University), Xia, Yong (Northwestern Polytechnical University)
SegmentationTransformerMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a multi-modal Siamese encoder network SMF-Net based on Swin Transformer for stroke lesion segmentation in multi-modal MRI (DWI, ADC, FLAIR) images.
SOFA: Deep Learning Framework for Simulating and Optimizing Atrial Fibrillation Ablation
Chung, Yunsung (Tulane University), Hamm, Jihun (Tulane University)
OptimizationMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Simulate the scars formed after atrial fibrillation ablation through deep learning and predict the risk of recurrence, further optimizing ablation parameters.
Solving Medical Multi-Label Domain Adaptation via Wasserstein Adversarial Learning with Class-Level Alignment
Liu, Wenjie (University of Science and Technology of China), Wang, Xu (University of Science and Technology of China)
Domain AdaptationConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical Data
🎯 What it does: This paper proposes the WAL-CLA method, which achieves medical multi-label domain adaptation through Wasserstein adversarial learning and class-level alignment.
Source-Free Active Domain Adaptation for Efficient Medical Video Polyp Segmentation
Li, Jialu (Hong Kong University of Science and Technology (Guangzhou)), Zhu, Lei (Hong Kong University of Science and Technology)
SegmentationDomain AdaptationTransformerVideoBiomedical Data
🎯 What it does: Proposes the first source-free active domain adaptation method for multi-center medical video data, aimed at colon polyp video segmentation.
Source-Free Domain Adaptation for Cross-Modality Cardiac Image Segmentation with Contrastive Class Relationship Consistency
Ma, Ao (Southwestern University of Finance and Economics), Chen, Xu (Southwestern University of Finance and Economics)
SegmentationDomain AdaptationConvolutional Neural NetworkContrastive LearningImageMultimodalityMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A source-free domain adaptation framework is proposed, utilizing a teacher-student model and contrastive learning to improve cross-modal cardiac image segmentation.
Sparse Reconstruction of Optical Doppler Tomography with Alternative State Space Model and Attention
Li, Zhenghong (Stony Brook University), Ling, Haibin (Stony Brook University)
RestorationSuper ResolutionConvolutional Neural NetworkBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a Sparse Optical Doppler Tomography Reconstruction Network based on an Alternative State Space Model and Attention Mechanism (ASSAN), which can quickly generate high-quality B-Scan images with fewer A-Scan data.
Sparse-XM: Spine Pose Adjustment with RGB-D Bone Segmentation via Cross-Modality Label Transfer
Warner, William R. (Dartmouth College), Fan, Xiaoyao (Dartmouth College)
SegmentationPose EstimationConvolutional Neural NetworkImagePoint CloudComputed Tomography
🎯 What it does: A fully automated framework for horizontal pose correction (Sparse-XM) is proposed, utilizing intraoperative stereo vision (iSV) images for spinal bone surface segmentation, and achieving spinal pose adjustment through alignment of the bone surface with preoperative CT (pCT).
Sparse3Diff: A Diffusion Framework for 3D Reconstruction from Sparse 2D Slices in Volumetric Optical Imaging
Lee, Hyun Jung (Korea University), Kam, Tae-Eui (Korea University)
RestorationGenerationDiffusion modelImageBiomedical Data
🎯 What it does: The Sparse3Diff framework is proposed to achieve high-fidelity 3D volume reconstruction from sparse 2D slices.
Sparsely Annotated Medical Image Segmentation via Cross-SAM of 3D and 2D Networks
Su, Huaqiang (Shenzhen University), Lei, Baiying (Shenzhen University)
SegmentationConvolutional Neural NetworkGraph Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: The SA-Net framework is proposed, utilizing the Segment Anything Model (SAM) with cross-supervision from 2D and 3D networks, combined with Bidirectional Graph Convolution (BGC), Multi-Scale Attention (MSA), and a Recovery Module (RM) to achieve sparse annotation in medical image segmentation.
Sparsely Labeled fMRI Data Denoising with Meta-Learning-Based Semi-Supervised Domain Adaptation
Heo, Keun-Soo (Korea University), Kam, Tae-Eui (Korea University)
RestorationDomain AdaptationMeta LearningConvolutional Neural NetworkContrastive LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A meta-learning based semi-supervised domain adaptation framework is proposed to learn noise features and perform denoising from sparsely labeled fMRI data.
Spatial Aggregation for Semi-supervised Active Learning in 3D Medical Image Segmentation
Ma, Siteng (University College Dublin), Dong, Ruihai (University College Dublin)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: The STAR framework is proposed, combining semi-supervised learning and active learning, generating pseudo-labels and queries between 2D slices of 3D medical images through spatial cross-attention, significantly reducing annotation costs.
Spatial Prior-Guided Boundary and Region-Aware 2D Lesion Segmentation in Neonatal Hypoxic Ischemic Encephalopathy
Rao, Amog (Plaksha University), Bao, Rina (Plaksha University)
SegmentationConvolutional Neural NetworkImageMagnetic Resonance Imaging
🎯 What it does: An efficient segmentation framework for neonatal hypoxic-ischemic encephalopathy (HIE) lesions based on a 2D UNet++ network is proposed, utilizing three-channel input (ADC, ZADC, and thresholded ZADC) combined with scSE attention and TLHF loss for precise segmentation.
Spatial regularisation for improved accuracy and interpretability in keypoint-based registration
Billot, Benjamin (Inria), Golland, Polina (Boston Children's Hospital)
SegmentationExplainability and InterpretabilityConvolutional Neural NetworkTime SeriesBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A triple spatial regularization loss is proposed, significantly improving the performance of unsupervised keypoint detection-based registration methods in terms of accuracy and interpretability.
Spatial-Temporal Memory Filtering SAM for Lesion Segmentation in Breast Ultrasound Videos
Tu, Zhengzheng (Anhui University), Zhang, Chaoxue (Anhui University)
SegmentationImageVideoUltrasound
🎯 What it does: A spatiotemporal memory filtering network based on the Segment Anything model (STMFSAM) is designed and implemented for lesion segmentation in breast ultrasound videos.
Spatially Gene Expression Prediction using Dual-Scale Contrastive Learning
Qu, Mingcheng (Harbin Institute of Technology), Fan, Lei (University of New South Wales)
SegmentationRepresentation LearningGraph Neural NetworkContrastive LearningImageMultimodalityBiomedical Data
🎯 What it does: A dual-scale contrastive learning framework NH²2ST is proposed, which predicts spatial gene expression from whole slide images by combining cross-attention, contrastive learning, and hypergraph learning in the query branch and neighborhood branch.
Spatio-temporal Pre-trained Foundation Model for Neural Decoding with Fine-grained Optimization
Li, Ziyu (Beijing Institute of Technology), Wu, Xia (Beijing Institute of Technology)
OptimizationGraph Neural NetworkSupervised Fine-TuningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Proposed and implemented a spatiotemporal pre-training foundational model based on fMRI, STPTF, achieving a neural decoding process with unsupervised data pre-training and fine-tuning with a small amount of labeled data.
Spatiotemporal-Sensitive Network for Microvascular Obstruction Segmentation from Cine Cardiac Magnetic Resonance
Yu, Yang (Institute for Infocomm Research, A*STAR), Yang, Xulei (Institute for Infocomm Research, A*STAR)
SegmentationConvolutional Neural NetworkVideoBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A spatiotemporal sensitive network is proposed, combining static and motion encoders, a guided decoder, and uncertainty-driven refinement to achieve segmentation of microvascular obstruction in non-contrast Cine CMR videos.
SPEC-CXR: Advancing Clinical Safety through Entity-Level Performance Evaluation of Chest X-ray Report Generation
Lee, Jung Oh (Lunit Inc.), Kim, Taesoo (Lunit Inc.)
GenerationSafty and PrivacyTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringTextBiomedical DataElectronic Health Records
🎯 What it does: A safety assessment framework SPEC-CXR for generating chest X-ray reports is proposed, which combines large language models for entity-level and report-level structured evaluation of generated reports against reference reports.
Speech Audio Generation from Dynamic MRI via a Knowledge Enhanced Conditional Variational Autoencoder
Li, Yaxuan (University of Hong Kong), Xing, Fangxu (Harvard Medical School)
GenerationData SynthesisConvolutional Neural NetworkTransformerAuto EncoderGenerative Adversarial NetworkBiomedical DataMagnetic Resonance ImagingAudio
🎯 What it does: This paper proposes a knowledge-enhanced conditional variational autoencoder (KE-CVAE) for directly generating speech waveforms from dynamic MRI sequences, addressing issues of audio noise, distortion, and synchronization in the MRI acquisition environment.
SPENet: Self-guided Prototype Enhancement Network for Few-shot Medical Image Segmentation
Fan, Chao, Wang, Liang (Capital Medical University Affiliated Beijing Friendship Hospital)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes the Self-guided Prototype Enhancement Network (SPENet), which addresses the intra-class variation problem in few-shot medical image segmentation through multi-layer prototype generation and query-guided local prototype enhancement.
Spherical Diffusion Process for Score-Guided Cortical Correspondence via Spectral Attention
Lee, Seungeun (Klleon), Lyu, Ilwoo (POSTECH)
Diffusion modelScore-based ModelMeshBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A framework for unsupervised cortical shape correspondence based on spherical diffusion processes and spectral attention mechanisms is designed, utilizing spherical harmonic space to learn conditional score functions to guide spherical deformation.
Spline refinement with differentiable rendering
Zdyb, Frans (University of Copenhagen), Kirkegaard, Julius B. (University of Copenhagen)
SegmentationOptimizationImage
🎯 What it does: A novel unsupervised differentiable rendering spline refinement method is proposed, which can achieve pixel-level accuracy of centerlines from coarse predictions.
SPromptGL: Semantic Prompt Guided Graph Learning for Multi-modal Brain Disease
Wan, Xixi (Anhui University), Zheng, Aihua (Anhui University)
ClassificationGraph Neural NetworkPrompt EngineeringMultimodalityBiomedical DataAlzheimer's Disease
🎯 What it does: A multi-modal brain disease prediction framework based on multi-relation graph learning and semantic prompting (SPromptGL) is proposed, which constructs a multi-modal relation graph, graph convolutional network, and prompt-guided embedding network to mine the discriminative lesion areas of each modality and fuse global features.
SpurBreast: A Curated Dataset for Investigating Spurious Correlations in Real-world Breast MRI Classification
Won, Jong Bum, Ozbulak, Utku (Ghent University Global Campus)
ClassificationConvolutional Neural NetworkTransformerSupervised Fine-TuningImageBiomedical DataMagnetic Resonance ImagingBenchmark
🎯 What it does: Constructed the SpurBreast dataset, intentionally introducing controllable pseudo-correlated features in real breast MRI images to systematically study their impact on the performance and generalization of deep learning models.
SR-SAM: Subspace Regularization for Domain Generalization of Segment Anything Model
Jiang, Xixi (Hong Kong University of Science and Technology), Yang, Xin (Shenzhen University)
SegmentationDomain AdaptationKnowledge DistillationConvolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes SR-SAM, which utilizes subspace regularization and EMA LoRA for parameter-efficient fine-tuning of the Segment Anything Model to enhance the cross-domain generalization ability of single-source medical image segmentation.
SSPNet: Towards Feasible Spatio-Spectral Portraits-Based Deep Learning Framework for Neurodegenerative Disease Multi-Classification
Kim, Ho-Jung (Hallym University), Won, Dong-Ok (Hallym University)
ClassificationConvolutional Neural NetworkTime SeriesBiomedical DataAlzheimer's Disease
🎯 What it does: A deep learning framework called SSPNet based on EEG spatiotemporal spectral portraits has been developed for the classification of multiple categories of neurodegenerative diseases (AD, FTD, normal).
STEAM: Self-supervised TEeth Analysis and Modeling for Point Cloud Segmentation
Liu, Yifan (Chinese University of Hong Kong), Yuan, Yixuan (Southern University of Science and Technology)
SegmentationTransformerAuto EncoderPoint Cloud
🎯 What it does: For the segmentation of dental point clouds from intraoral scans, a self-supervised STEAM framework is proposed, which completes segmentation after pre-training on a large amount of unlabeled point clouds followed by fine-tuning.
Steerable Anatomical Shape Synthesis with Implicit Neural Representations
de Wilde, Bram (University of Twente), Wolterink, Jelmer M. (University of Twente)
SegmentationGenerationData SynthesisImageBiomedical DataComputed Tomography
🎯 What it does: A guided generative model based on implicit neural representations is proposed for synthesizing anatomical structures, particularly the thyroid, which can be controlled for specific patient populations.
StepAL: Step-aware Active Learning for Cataract Surgical Videos
Shah, Nisarg A. (Johns Hopkins University), Patel, Vishal M. (Johns Hopkins University)
RecognitionTransformerVideo
🎯 What it does: A proactive learning framework StepAL for recognizing surgical steps in long videos is designed, selecting labeled samples based on the overall video.
STMDiff: Spatiotemporal Matching Diffusion Model for Dual-Time-Point Total-body PET/CT Imaging via Contrastive Learning
Li, Wenbo (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences), Hu, Zhanli (United Imaging Healthcare Group)
RestorationGenerationDiffusion modelContrastive LearningImageBiomedical DataComputed TomographyPositron Emission Tomography
🎯 What it does: Utilizing the CT images from the first time point and the PET images from the second time point for spatiotemporal matching, and generating CT-free attenuation-corrected PET images based on a diffusion model, thereby eliminating the radiation risk associated with the second CT scan in dual-time-point PET/CT scans.
Stronger Together: Registering Preoperative Imagery, LUS, and MIS Liver Images
Kalantari, Mohammad Mahdi (Clermont Auvergne INP), Bartoli, Adrien (University Hospital of Clermont Ferrand)
SegmentationOptimizationImageMultimodalityMagnetic Resonance ImagingComputed TomographyUltrasound
🎯 What it does: This paper proposes a multimodal registration method that does not require external sensors or markers, combining preoperative 3D models, laparoscopic ultrasound (LUS), and intraoperative laparoscopic images to achieve precise localization of liver tumors.
Structure and Smoothness Constrained Dual Networks for MR Bias Field Correction
Liang, Dong (Chinese Academy of Sciences), Luo, Gongning (King Abdullah University of Science and Technology)
RestorationSegmentationConvolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A method for MR bias field correction based on a self-supervised dual network with structural and smoothness constraints (S2DNets) is proposed.
Structure-Aware Cross-Modal Prompt Tuning for Autonomous Bronchoscopic Navigation
Fang, Hao (Xiamen University), Luo, Xiongbiao (Xiamen University)
RecognitionSegmentationAutonomous DrivingTransformerPrompt EngineeringContrastive LearningImageMultimodality
🎯 What it does: This paper proposes a structure-aware cross-modal prompt tuning framework (SCPT), which enhances the fine-grained recognition of bronchial bifurcations and open-set recognition performance by integrating CLIP with EfficientSAM and combining cross-modal attention with a base-unknown knowledge decoupling head.
Structure-aware MRI Translation: Multi-Modal Latent Diffusion Model with Arbitrary Missing Modalities
Zhang, Xinzhe (Northeastern University), Zaiane, Osmar R. (Amii)
Image TranslationGenerationData SynthesisDiffusion modelMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A multimodal MRI missing modality translation framework named MISA-LDM is proposed, which can synthesize the missing MRI modality under any available modality combination and ensures the structural consistency of the generated images through a structure-preserving mechanism.
Structure-Preserve Expansion for Medical Image Registration with Minimal Overlap
Liu, Zaiyuan (Northwestern Polytechnical University), Xia, Yong (Northwestern Polytechnical University)
Image TranslationGenerationOptimizationConvolutional Neural NetworkTransformerGenerative Adversarial NetworkImageBiomedical DataComputed Tomography
🎯 What it does: A Sequential Structure Preserving Extension (SSPE) framework is proposed, which uses a generative network to gradually expand partially overlapping medical images, thereby increasing the overlapping area and improving the registration performance of images with minimal overlap.
StyleGAN-based Brain MRI Anomaly Detection via Latent Code Retrieval and Partial Swap
Wei, Jie, Wang, Guotai (University Of Electronic Science And Technology Of China)
Anomaly DetectionGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A method for detecting abnormalities in brain MRI based on StyleGAN is proposed, constructing a latent code library of healthy images, using coarse-to-fine retrieval and local latent code exchange to reconstruct normal images and locate abnormalities.
Subgroup Performance Analysis in Hidden Stratifications
Bissoto, Alceu (University of Bern), Koch, Lisa M. (University of Bern)
ClassificationAnomaly DetectionContrastive LearningImageBiomedical Data
🎯 What it does: This paper utilizes subgroup discovery techniques for performance analysis in medical image classification, exploring the model performance differences caused by hidden stratification.
Subtyping Breast Lesions via Generative Augmentation based Long-tailed Recognition in Ultrasound
Chen, Shijing (Shenzhen University), Huang, Ruobing (Shenzhen University)
ClassificationRecognitionReinforcement LearningDiffusion modelImageBiomedical DataUltrasound
🎯 What it does: A dual-stage framework is proposed to achieve long-tail recognition of breast ultrasound image subtypes through high-fidelity synthesis and reinforcement learning adaptive sampling.
SUGFW: A SAM-based Uncertainty-guided Feature Weighting Framework for Cold Start Active Learning
Ma, Xiaochuan (University of Electronic Science and Technology of China), Wang, Guotai (University Of Electronic Science And Technology Of China)
SegmentationImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A cold-start active learning framework SUGFW based on the Segment Anything Model (SAM) is proposed, which utilizes SAM's zero-shot segmentation capability to estimate uncertainty. It achieves feature weighting and efficient sample selection through Patch-based Global Distinct Representation (PGDR) and Greedy Selection with Cluster and Uncertainty (GSCU) to enhance the labeling efficiency of medical image segmentation.
Super-resolution and segmentation of 4D Flow MRI using Deep learning and Weighted Mean Frequencies
Perrin, Simon, Serfaty, Jean-Michel
SegmentationSuper ResolutionBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The content of this paper is unidentifiable.
Supervised Diffusion-Model-Based PET Image Reconstruction
Webber, George (King's College London), Reader, Andrew J. (King's College London)
RestorationDiffusion modelScore-based ModelImageBiomedical DataPositron Emission TomographyStochastic Differential Equation
🎯 What it does: A supervised diffusion model (DM) framework called PET-DEFT is proposed for PET image reconstruction, balancing data consistency and prior learning, and supporting posterior sampling.
Surface Vision Mamba: Leveraging Bidirectional State Space Model for Efficient Spherical Manifold Representation
He, Rongzhao (Lanzhou University), Hu, Bin (Lanzhou University)
Computational EfficiencyRepresentation LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper presents Surface Vision Mamba (SiM), an efficient visual model for spherical surface data aimed at predicting neonatal brain development phenotypes.
Surface-based Multi-Axis Longitudinal Disentanglement Using Contrastive Learning for Alzheimer’s Disease
Zhang, Jianwei (University of Southern California), Shi, Yonggang (University of Southern California)
ClassificationExplainability and InterpretabilityRepresentation LearningGraph Neural NetworkAuto EncoderContrastive LearningImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A multi-axis longitudinal decoupling framework based on cortical surface maps is proposed, utilizing self-supervised contrastive learning to separate aging effects from multiple Alzheimer's disease trajectories, forming an interpretable latent space.
Surgical Action Planning with Large Language Models
Xu, Mengya (Chinese University of Hong Kong), Dou, Qi (Huazhong University of Science and Technology)
TransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringVision Language ModelVideoTextMultimodality
🎯 What it does: This paper proposes a surgical action planning (SAP) task for cholecystectomy and constructs the CholecT50-SAP dataset, utilizing large language models to generate future surgical actions and interpretable text outputs.
Surgical-MambaLLM: Mamba2-enhanced Multimodal Large Language Model for VQLA in Robotic Surgery
Hao, Pengfei (Hong Kong University of Science and Technology), Zhu, Lei (Hong Kong University of Science and Technology)
Robotic IntelligenceTransformerLarge Language ModelSupervised Fine-TuningVision Language ModelImageMultimodality
🎯 What it does: This paper proposes the Surgical-MambaLLM model, which combines Mamba2 with large language models for the surgical visual question answering localization answering (VQLA) task.
SurgicalGS: Dynamic 3D Gaussian Splatting for Accurate Robotic-Assisted Surgical Scene Reconstruction
Chen, Jialei (Imperial College London), Huang, Baoru (University College London)
Depth EstimationRobotic IntelligenceGaussian SplattingPoint Cloud
🎯 What it does: The SurgicalGS framework is proposed to achieve high-precision 3D reconstruction of dynamic surgical scenes through multi-frame depth fusion, adaptive motion masks, and flexible deformation models.
SurgSora: Object-Aware Diffusion Model for Controllable Surgical Video Generation
Chen, Tong (University of Sydney), Zhou, Luping (University of Sydney)
GenerationData SynthesisDiffusion modelVideo
🎯 What it does: This paper presents SurgSora, a controllable single-frame surgical video generation method based on diffusion models, allowing users to control the movement of tools and tissues through trajectory clicks.
SurgTPGS: Semantic 3D Surgical Scene Understanding with Text Promptable Gaussian Splatting
Huang, Yiming (Chinese University of Hong Kong), Ren, Hongliang (Chinese University of Hong Kong)
Object DetectionSegmentationConvolutional Neural NetworkVision Language ModelAuto EncoderGaussian SplattingVideo
🎯 What it does: The SurgTPGS method is proposed, enabling 3D surgical scene semantic segmentation and querying through text prompts.
SurgX: Neuron-Concept Association for Explainable Surgical Phase Recognition
Kim, Ka Young (Kyung Hee University), Kim, Seong Tae (Technical University of Munich)
RecognitionExplainability and InterpretabilityTransformerVision Language ModelVideo
🎯 What it does: Proposes the SurgX framework, which provides conceptual explanations for the neurons in the surgical video stage recognition model, constructs a dedicated concept set, and enhances explanation quality through representative sequences.
SV-DRR: High-Fidelity Novel View X-Ray Synthesis Using Diffusion Model
Xie, Chun (University of Tsukuba), Kitahara, Itaru (University of Tsukuba)
GenerationData SynthesisTransformerDiffusion modelImageComputed Tomography
🎯 What it does: A synthetic method for generating multi-angle high-resolution projections from single-view X-ray images (SV-DRR) has been designed.
Synchronous Inhibition and Activation for Weakly Supervised Semantic Segmentation of Pathology Images
Fan, Jiansong (Jiangnan University), Pan, Xiang (Hangzhou Dianzi University)
SegmentationGraph Neural NetworkTransformerContrastive LearningImage
🎯 What it does: A weakly supervised semantic segmentation model based on synchronous suppression and activation (SIA-WSSS) is proposed for pathological images.
Synergy-Guided Regional Supervision of Pseudo Labels for Semi-Supervised Medical Image Segmentation
Wang, Tao (Fuzhou University), Tong, Tong (Fuzhou University)
SegmentationConvolutional Neural NetworkBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes a collaborative guidance-based regional supervision framework SGRS-Net for semi-supervised medical image segmentation, enhancing the utilization of pseudo-labels.
SynPo: Boosting Training-Free Few-Shot Medical Segmentation via High-Quality Negative Prompts
Liu, Yufei, Luo, Zhiming (Peking University)
SegmentationTransformerContrastive LearningImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes a training-free few-shot medical segmentation method called SynPo, which generates high-quality positive and negative prompt points by constructing a collaborative confidence map that integrates DINOv2 and SAM, achieving precise segmentation.
Synthesis of Pathological Dual-Channel Color Doppler Echocardiograms for Equitable Diagnosis of Heart Diseases
Roshanitabrizi, Pooneh (Children's National Hospital), Linguraru, Marius George (George Washington University)
ClassificationGenerationData SynthesisConvolutional Neural NetworkDiffusion modelImageVideoBiomedical DataUltrasound
🎯 What it does: A 3.5D conditional diffusion model is proposed to synthesize dual-channel color Doppler cardiac ultrasound and B-mode images, and synthetic data is used to enhance RHD detection performance.
Synthesizing Delayed-Phase Contrast-Enhanced Breast MR Images from Early-Phase Images Using an Iterative Deep Network
Chung, Woojin (Hankuk University of Foreign Studies), Nam, Yoonho (Hankuk University of Foreign Studies)
GenerationData SynthesisConvolutional Neural NetworkRecurrent Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An iterative deep network, iterU-Net, is proposed to synthesize multi-time point delayed phase-enhanced images from early dynamic contrast-enhanced breast MRI images, achieving dynamic enhancement curve prediction.
Synthetic Ground Truth Counterfactuals for Comprehensive Evaluation of Causal Generative Models in Medical Imaging
Stanley, Emma A. M. (University of Calgary), Wilms, Matthias (University of Calgary)
GenerationData SynthesisImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Using synthetic brain MRI ground truth counterfactual data generated by SimBA to evaluate and debug the effects of causal generative models in medical imaging.
T&F-DFC FusionNet: Time&Frequency-Dynamic Functional Connectivity Fusion Network for ADHD Diagnosis in Children based on fNIRS
Chu, Mengxiang (Northwest University), Guo, Hongbo (Northwest University)
ClassificationConvolutional Neural NetworkRecurrent Neural NetworkTime SeriesBiomedical Data
🎯 What it does: This paper proposes a time-frequency dynamic functional connectivity fusion network (T&F-DFC FusionNet) based on fNIRS for the objective diagnosis of ADHD in children.
T2GS: Comprehensive Reconstruction of Dynamic Surgical Scenes with Gaussian Splatting
Xu, Jinjing (National Center for Tumor Diseases), Speidel, Stefanie (National Center for Tumor Diseases)
RestorationObject TrackingPose EstimationOptimizationGaussian SplattingVideoPoint Cloud
🎯 What it does: Using 3D Gaussian splatting to construct the T²GS framework, a unified and efficient three-dimensional reconstruction of soft tissue deformation and surgical tool movement has been achieved.
T2I-Diff: fMRI Signal Generation via Time-Frequency Image Transform and Classifier-Free Denoising Diffusion Models
Tew, Hwa Hui (Monash University Malaysia), Ting, Chee-Ming (King Abdullah University of Science and Technology)
ClassificationGenerationData SynthesisDiffusion modelTime SeriesBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Proposes the T2I-Diff framework, which utilizes time-frequency image representation and a classifier-free denoising diffusion model to generate fMRI BOLD signals for brain network classification.
T2WI-BCMIC: Non-Fat Saturated T2-Weighted Imaging Dataset for Bladder Cancer Muscle Invasion Classification
Huang, Han (South China University of Technology), Guo, Yan (First Affiliated Hospital Sun Yat Sen University)
ClassificationOptimizationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingBenchmark
🎯 What it does: The first publicly available multi-class dataset for the risk of muscle invasion in bladder cancer (T2WI-BCMIC) has been proposed and released, and a benchmark model has been constructed based on this dataset. Subsequently, a constraint optimization-based search data augmentation algorithm has been proposed to enhance image classification performance.
Tackling Hallucination from Conditional Models for Medical Image Reconstruction with DynamicDPS
Kim, Seunghoi (University College London), Alexander, Daniel C.
RestorationDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes the DynamicDPS framework, which combines unconditional diffusion models with data consistency. It first uses a conditional model to generate high-field predictions of low-field MRI, and then suppresses artifacts through dynamic diffusion posterior sampling, achieving high-quality reconstruction of low-field MRI.
Target Prior-enriched Implicit 3D CT Reconstruction with Adaptive Ray Sampling
Cao, Qinglei (Southwest University), Tang, Xiaoqin (Southwest University)
RestorationOptimizationComputational EfficiencyNeural Radiance FieldImageBiomedical DataComputed Tomography
🎯 What it does: A target prior-based implicit 3D CT reconstruction framework TP-INR is proposed, which utilizes sparse projections to generate anatomical priors and integrates structural and positional information to achieve high-quality reconstruction.
Targeted False Positive Synthesis via Detector-guided Adversarial Diffusion Attacker for Robust Polyp Detection
Zhou, Quan (Wuhan University of Technology), Wang, Zhiwei (Huazhong University of Science and Technology)
Object DetectionAdversarial AttackDiffusion modelImage
🎯 What it does: A detector-guided adversarial diffusion attack framework (DADA) is proposed, which trains a diffusion model that generates only backgrounds and injects adversarial perturbations during the diffusion process to synthesize high-value false positive samples to enhance polyp detection accuracy.