MICCAI 2025 Papers — Page 10
International Conference on Medical Image Computing and Computer-Assisted Intervention · 1027 papers
Task-aligned fMRI Generation Model for Brain Disorder Diagnosis
Li, Yifan (Sun Yat-sen University), Zhang, Jianjia (Sun Yat-sen University)
ClassificationGenerationData SynthesisDiffusion modelTime SeriesBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: Generate fMRI time series data through a diffusion model combined with functional brain networks (FBN), incorporating label information during the generation process to align the generated fMRI with brain disease diagnosis tasks.
TAT: Task-Adaptive Transformer for All-in-One Medical Image Restoration
Yang, Zhiwen (Beihang University), Xu, Yan (Beihang University)
RestorationSuper ResolutionTransformerImageBiomedical DataMagnetic Resonance ImagingComputed TomographyPositron Emission Tomography
🎯 What it does: This paper proposes a Task-Adaptive Transformer (TAT) for integrated medical image restoration across multiple tasks (PET synthesis, CT denoising, MRI super-resolution);
tCURLoRA: Tensor CUR Decomposition Based Low-Rank Parameter Adaptation and Its Application in Medical Image Segmentation
He, Guanghua (Hangzhou Dianzi University), Yu, Gaohang (Hangzhou Dianzi University)
SegmentationTransformerBiomedical DataAlzheimer's Disease
🎯 What it does: This paper proposes a low-rank parameter adaptation method based on tensor CUR decomposition, called tCURLoRA, for efficient fine-tuning in medical image segmentation.
Teaching pathology foundation models to accurately predict gene expression with parameter efficient knowledge transfer
Pan, Shi (University College London), Secrier, Maria (University College London)
Domain AdaptationKnowledge DistillationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: The PEKA framework is proposed to perform cross-modal knowledge transfer between the pathology image base model and the single-cell transcriptome base model to enhance the accuracy of gene expression prediction.
TEGDA: Test-time Evaluation-Guided Dynamic Adaptation for Medical Image Segmentation
Zhou, Yubo (University of Electronic Science and Technology of China), Wang, Guotai (University Of Electronic Science And Technology Of China)
SegmentationDomain AdaptationSupervised Fine-TuningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A dynamic adaptation framework TEGDA based on test-time evaluation is proposed for test-time adaptation in medical image segmentation across different domains.
Temporal Atlas-Guided Generation of Longitudinal Data via Geometric Latent Embeddings
Wu, Shaoju (Boston Children's Hospital and Harvard Medical School), Tsai, Andy (Boston Children's Hospital and Harvard Medical School)
GenerationData SynthesisImageBiomedical DataComputed Tomography
🎯 What it does: A deep learning framework based on a temporal atlas is proposed to generate individualized longitudinal (cross-age) anatomical morphology sequences from simple cross-sectional CT data.
Temporal Differential Fields for 4D Motion Modeling via Image-to-Video Synthesis
You, Xin (Shanghai Jiao Tong University), Navab, Nassir (Technische Universität München)
GenerationData SynthesisDiffusion modelImageVideoBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A diffusion framework Mo-Diff based on image-to-video (I2V) is proposed, which predicts future four-dimensional medical image sequences from a single frame.
Temporal Model-Based Federated Active Medical Image Classification
Yan, Yunlu (Hong Kong University of Science and Technology (Guangzhou)), Zhu, Lei (Hong Kong University of Science and Technology)
ClassificationFederated LearningImageBiomedical Data
🎯 What it does: A federated active learning framework for multi-institutional medical image classification, TM-FAL, is proposed, which utilizes temporal local and global model estimates of sample uncertainty and mitigates class imbalance through pseudo-label grouping.
Temporal Modulated Multi-Scale Deformation Fusion via Knowledge Distillation for 4D Medical Image Interpolation
Zhang, Jiaju (Bejing Institute of Technology), Yang, Jian (Bejing Institute of Technology)
Image TranslationRestorationKnowledge DistillationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes a multi-scale time-varying deformation fusion framework based on knowledge distillation, which can directly predict the deformation field at any intermediate time point and generate interpolated volumes given only the two end time phases of 4D medical images.
Temporal Neural Cellular Automata: Application to modeling of contrast enhancement in breast MRI
Lang, Daniel M. (Helmholtz Munich), Schnabel, Julia A. (Technical University of Munich)
RestorationGenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A time-based neural cellular automata (TeNCA) model is used to model dynamic contrast enhancement in breast MRI and achieve synthetic prediction for time-sparse, non-uniformly sampled sequences.
Temporal Representation Learning of Phenotype Trajectories for pCR Prediction in Breast Cancer
Janíčková, Ivana (Medical University of Vienna), Langs, Georg (Medical University of Vienna)
ClassificationRepresentation LearningConvolutional Neural NetworkImageTime SeriesBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a multi-task self-supervised pre-training framework that learns representations from the temporal trajectories of breast MRI follow-up sequences and predicts whether pathological complete response (pCR) is achieved after axillary neoadjuvant chemotherapy.
Temporally-Aware Supervised Contrastive Learning for Polyp Counting in Colonoscopy
Parolari, Luca (University of Padova), Biffi, Carlo (Cosmo Intelligent Medical Devices)
Object DetectionSegmentationRepresentation LearningTransformerContrastive LearningVideoBiomedical Data
🎯 What it does: This paper proposes a temporal-aware supervised contrastive learning framework to improve the accuracy of polyp counting in colonoscopy videos.
TemSAM: Temporal-aware Segment Anything Model for Cerebrovascular Segmentation in Digital Subtraction Angiography Sequences
Zhang, Liang (Huazhong University of Science and Technology), Yang, Xin (Shenzhen University)
SegmentationImageVideoBiomedical Data
🎯 What it does: This paper proposes TemSAM, a time-aware segmentation model for digital subtraction angiography sequences, which significantly improves vascular segmentation accuracy by combining MIP global cues and complementary frame fusion.
TESLA: Test-time Reference-free Through-plane Super-resolution for Multi-contrast Brain MRI
Choi, Yoonseok (Yonsei University), Kim, Dong-Hyun (Yonsei University)
RestorationSuper ResolutionContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A two-stage network called TESLA is proposed, which achieves multi-contrast brain MRI through-plane super-resolution using progressive reconstruction and structural enhancement, and does not require reference images during inference.
Test-Time Training with Local Contrast-Preserving Copy-Pasted Image for Domain Generalization in Retinal Vessel Segmentation
Gu, Yuliang (Wuhan University), Xu, Yongchao (Wuhan University)
SegmentationDomain AdaptationConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: A test-time training method for domain generalization in retinal vessel segmentation is proposed, which generates synthetic images through local contrast-preserving copy-paste (L2CP) during testing.
Tetra-orientated Mamba with T2-FLAIR Mismatch Features for Glioma Segmentation, IDH Genotyping, and Grading
Li, Xinyu (Central South University), Wang, Jianxin (Institute of Guizhou Aerospace Measuring and Testing Technology)
ClassificationSegmentationTransformerMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The MTamba multi-task network is proposed to achieve three-dimensional segmentation of brain gliomas, IDH gene typing, and joint prediction of grading.
Text-driven Multiplanar Visual Interaction for Semi-supervised Medical Image Segmentation
Huang, Kaiwen (Nanjing University of Science and Technology), Zhou, Tao (Nanjing University of Science and Technology)
SegmentationPrompt EngineeringContrastive LearningImageBiomedical DataComputed Tomography
🎯 What it does: A Text-SemiSeg framework is proposed, which combines text information with 3D medical image segmentation to achieve semi-supervised learning.
Text-Guided Multi-Instance Learning for Scoliosis Screening via Gait Video Analysis
Li, Haiqing (University of Texas at Arlington), Huang, Junzhou (University of Texas at Arlington)
ClassificationRecognitionTransformerLarge Language ModelVideoText
🎯 What it does: This paper proposes a text-guided multi-instance learning network (TG-MILNet) for detecting early scoliosis in adolescents through gait videos.
TextBraTS: Text-Guided Volumetric Brain Tumor Segmentation with Innovative Dataset Development and Fusion Module Exploration
Shi, Xiaoyu (Ritsumeikan University), Chen, Yen-wei (Ritsumeikan University)
SegmentationTransformerLarge Language ModelImageTextMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study created the first volumetric text-image brain tumor segmentation dataset, TextBraTS, and proposed a 3D brain tumor segmentation framework that utilizes sequential cross-attention to integrate text information based on this dataset.
TGSAM-2: Text-Guided Medical Image Segmentation using Segment Anything Model 2
Yuan, Runtian (Fudan University), Gao, Shang (Shanghai University of Finance and Economics)
SegmentationTransformerPrompt EngineeringImageVideoBiomedical DataMagnetic Resonance ImagingComputed TomographyUltrasound
🎯 What it does: In the Segment Anything Model 2 (SAM-2) framework, text prompts are incorporated to propose TGSAM-2, achieving medical image segmentation guided by medical text.
The Missing Piece: A Case for Pre-Training in 3D Medical Object Detection
Eckstein, Katharina (German Cancer Research Center), Maier-Hein, Klaus H. (German Cancer Research Center)
Object DetectionSupervised Fine-TuningContrastive LearningBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: The system evaluates the impact of large-scale pre-training on 3D medical object detection, comparing supervised and various self-supervised methods, and verifies their enhancement effects on different detection architectures.
The Refining of Brain Connectivity Features on Residual Posterior Patterns
Zha, Xinbei, Gu, Jin (Southwest Jiaotong University)
ClassificationAnomaly DetectionGraph Neural NetworkGraphBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a Residual-Posterior Line Graph Network (RP-LGN) that transforms brain functional connections (edges) into nodes for learning, enhancing the diagnostic capability for neuropsychiatric disorders such as autism and ADHD.
Think as Cardiac Sonographers: Marrying SAM with Left Ventricular Indicators Measurements According to Clinical Guidelines
Liu, Tuo (Southeast University), Zhou, Guangquan (Southeast University)
Convolutional Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: The AutoSAME framework is proposed to achieve automated measurement of left ventricular indicators in cardiac ultrasound based on SAM.
This EEG Looks Like These EEGs: Interpretable Interictal Epileptiform Discharge Detection With ProtoEEG-kNN
Tang, Dennis (Duke University), Rudin, Cynthia (Harvard Medical School)
ClassificationExplainability and InterpretabilityConvolutional Neural NetworkTime SeriesBiomedical Data
🎯 What it does: This paper proposes the ProtoEEG-kNN model for interpretable detection of interictal epileptiform discharges (IEDs) and provides diagnostic explanations through case comparisons.
tHPM-LDM: Integrating Individual Historical Record with Population Memory in Latent Diffusion-based Glaucoma Forecasting
Fan, Yuheng (University of Liverpool), Zhao, He (Liverpool University Hospitals Nhs Foundation Trust)
ClassificationGenerationTransformerDiffusion modelContrastive LearningImageTime SeriesBiomedical Data
🎯 What it does: A glaucoma prediction framework based on a conditional latent diffusion model (t HPM‑LDM) is proposed, which captures the evolution of individual irregular follow-up records using continuous-time attention and introduces population memory modules to incorporate group progression information, enabling predictions of future retinal images and disease classification.
Thread the Needle: Genomics-guided Prompt-bridged Attention Model for Survival Prediction of Glioma based on MRI Images
Zhong, Yi (Harbin Institute of Technology), Zhang, Zhiguo (Harbin Institute of Technology)
ClassificationTransformerPrompt EngineeringImageMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A gene-guided prompt learning framework is proposed, utilizing multimodal fusion of MRI images and transcriptomic data to enhance the accuracy of survival prediction for brain gliomas.
ThyroidXL: Advancing Thyroid Nodule Diagnosis with an Expert-Labeled, Pathology-Validated Dataset
Duong , Viet Hung, Ngo, Dien Hy
Biomedical DataMagnetic Resonance Imaging
🎯 What it does: Unable to identify the content of the paper
Tied Prototype Model for Few-Shot Medical Image Segmentation
Kim, Hyeongji (UiT Arctic University of Norway), Kampffmeyer, Michael (UiT Arctic University of Norway)
SegmentationImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: Proposes the Tied Prototype Model (TPM), which improves ADNet for few-shot medical image segmentation, supports multiple prototypes and multiple categories, and introduces ideal threshold estimation;
Time-Contrastive Pretraining for In-Context Image and Video Segmentation
Wahd, Assefa (University of Alberta), Hareendranathan, Abhilash (University of Alberta)
SegmentationConvolutional Neural NetworkContrastive LearningImageVideoComputed TomographyBenchmark
🎯 What it does: A visual in-context learning framework called Temporal is proposed, based on video object segmentation (VOS), and a context retriever is pre-trained through temporal contrastive learning.
Time-Lapse Video-Based Embryo Grading via Complementary Spatial-Temporal Pattern Mining
Sun, Yong (Hong Kong University of Science and Technology (Guangzhou)), Nie, Qiang (Hong Kong University of Science and Technology (Guangzhou))
ClassificationRecognitionTransformerMixture of ExpertsVideo
🎯 What it does: A time-lapse video-based embryo grading framework called CoSTeM has been designed and implemented to automatically predict the overall quality grading of artificially fertilized embryos.
TMSE: Tri-Modal Survival Estimation with Context-aware Tissue Prototype and Attention-Entropy Interaction
Zhang, Ruofan (University of Chinese Academy of Sciences), Dong, Di (University of Chinese Academy of Sciences)
ClassificationData-Centric LearningLarge Language ModelMultimodalityBiomedical Data
🎯 What it does: A tri-modal survival prediction framework TMSE is proposed, integrating whole slide images (WSI), pathology reports, and genomic data. This is achieved through specialized report preprocessing, a Context-aware Tissue Prototype (CTP) module, and an Attention-Entropy Interaction (AEI) module to predict patient prognosis time.
Top-Down Attention-based Multiple Instance Learning for Whole Slide Image Analysis
Reisenbüchler, Daniel (University of Regensburg), Merhof, Dorit (Hannover Medical School)
ClassificationTransformerImage
🎯 What it does: A two-stage multi-instance learning framework TDA-MIL is proposed, combining self-attention and task-specific feature selection to achieve efficient classification of Whole Slide Images.
Topology-Constrained Learning for Efficient Laparoscopic Liver Landmark Detection
Cui, Ruize, Qin, Jing (Hong Kong Polytechnic University)
Object DetectionSegmentationComputational EfficiencyConvolutional Neural NetworkImageMultimodalityMagnetic Resonance Imaging
🎯 What it does: This paper proposes a lightweight topological constraint learning framework called TopoNet, aimed at achieving high precision and efficiency in landmark detection during laparoscopic liver surgery.
Toward Medical Deepfake Detection: A Comprehensive Dataset and Novel Method
Li, Shuaibo (Hong Kong University of Science and Technology), Zhu, Lei (Hong Kong University of Science and Technology)
RetrievalAnomaly DetectionConvolutional Neural NetworkVision Language ModelImageBiomedical DataMagnetic Resonance ImagingComputed TomographyUltrasoundRetrieval-Augmented Generation
🎯 What it does: This paper presents the MedForensics large dataset for medical deepfake detection and the Dual-Stage Knowledge Injection detection framework (DSKI) for identifying AI-generated medical images.
Towards Accurate Tumor Budding Detection: A Benchmark Dataset and A Detection Approach Based on Implicit Annotation Standardization and Positive-Negative Feature Coupling
Sun, Rui-Qing (Beijing Institute Of Technology), Zhang, Shaohui (Beijing Institute Of Technology)
Object DetectionKnowledge DistillationConvolutional Neural NetworkContrastive LearningImageBiomedical DataBenchmark
🎯 What it does: Developed a Tumor Budding Detection Network (TBDNet) based on YOLOv5 and released the first publicly available tumor budding detection benchmark dataset, TBDD.
Towards Automated Pediatric Dental Development Staging: A Dataset and Model
Wang, Peng (Nankai University), Li, Tao (Nankai University)
ClassificationRecognitionTransformerMixture of ExpertsImage
🎯 What it does: Created and released the DentalDS dataset for children's tooth development stages, and proposed DDSNet for tooth stage recognition.
Towards Globally Predictable k-Space Interpolation: A White-box Transformer Approach
Luo, Chen (Inner Mongolia University), Liang, Dong (Chinese Academy of Sciences)
TransformerBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Proposed GPI-WT white-box Transformer for k-space interpolation
Towards Holistic Surgical Scene Graph
Shin, Jongmin (Samsung Medical Center), Oh, Namkee (Samsung Medical Center)
RecognitionObject DetectionGraph Neural NetworkSupervised Fine-TuningVideo
🎯 What it does: This paper presents a new surgical scene graph dataset, Endoscapes‑SG201, and proposes a novel scene graph generation model, SSG‑Com, which can simultaneously capture the tool-action-target triplet relationships and hand identity information to enhance the understanding of surgical scenes.
Towards Interpretable Counterfactual Generation via Multimodal Autoregression
Ma, Chenglong (Fudan University), Shan, Hongming (Fudan University)
GenerationExplainability and InterpretabilityTransformerLarge Language ModelImageTextMultimodalityBiomedical DataElectronic Health Records
🎯 What it does: Proposes an explainable causal counterfactual generation task that jointly generates chest X-ray images and corresponding explanatory text, and releases the ICG-CXR dataset.
Towards Markerless Intraoperative Tracking of Deformable Spine Tissue
Daly, Connor (Imperial College London), Rodriguez y Baena, Ferdinando (Imperial College London)
Object TrackingSegmentationConvolutional Neural NetworkSupervised Fine-TuningPoint CloudMeshBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: The first real clinical RGB-D spinal surgery dataset is proposed and implemented, and based on this data, SpineAlign and CorrespondNet are developed to achieve markerless real-time tracking and registration of spinal soft tissues.
Towards Multi-Scenario Generalization: Text-Guided Unified Framework for Low-Dose CT and Total-Body PET Reconstruction
Wang, Weitao, Fu, Yu (Lanzhou University)
RestorationDiffusion modelImageMultimodalityBiomedical DataComputed TomographyPositron Emission Tomography
🎯 What it does: A text-guided unified framework (TUF) is proposed for high-precision reconstruction of low-dose CT and whole-body low-dose PET images.
Towards Patient-Specific Deformable Registration in Laparoscopic Surgery
Neri, Alberto (Istituto Italiano di Tecnologia), Mattos, Leonardo S. (Istituto Italiano di Tecnologia)
TransformerPoint CloudBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a non-rigid point cloud registration method based on Transformer, specifically trained for each patient, which can align the complete liver surface obtained from preoperative CT with the partial, deformed point cloud at the surgical site, enhancing the accuracy of surgical navigation.
Towards Robust Medical Image Referring Segmentation with Incomplete Textual Prompts
Wang, Qijie (Huazhong University of Science and Technology), Yan, Zengqiang (Huazhong University of Science and Technology)
SegmentationConvolutional Neural NetworkPrompt EngineeringImageTextBiomedical DataComputed Tomography
🎯 What it does: This study proposes a model named ARSeg for handling incomplete text prompts in the task of medical image segmentation.
Towards Robust Retinal Vessel Segmentation via Reducing Open-set Label Noises from SAM-generated Masks
Zhang, Minqing (Chinese University of Hong Kong), Yuan, Wu (Chinese University of Hong Kong)
SegmentationTransformerSupervised Fine-TuningImage
🎯 What it does: Utilizing finely annotated retinal vascular images to fine-tune the Segment Anything Model (SAM) to generate pseudo-labels, and eliminating open-set label noise through prototype recognition and context denoising, thereby achieving robust vascular segmentation.
TRACE: Temporally Reliable Anatomically-Conditioned 3D CT Generation with Enhanced Efficiency
Shao, Minye (Durham University), Zheng, Yefeng (Westlake University)
GenerationData SynthesisComputational EfficiencyDiffusion modelOptical FlowImageMultimodalityBiomedical DataComputed Tomography
🎯 What it does: Using a 2D multimodal diffusion model, 3D CT is treated as an ordered sequence of 2D slices to achieve 3D CT synthesis with variable axial lengths.
Training state-of-the-art pathology foundation models with orders of magnitude less data
Karasikov, Mikhail (kaiko.ai), Otálora, Sebastian (Netherlands Cancer Institute)
ClassificationSegmentationRepresentation LearningTransformerContrastive LearningImageBiomedical Data
🎯 What it does: Trained a small number (12k–92k) of WSI pathology visual foundation models, demonstrating that high performance can be achieved with less data.
Training-free Test-time Improvement for Explainable Medical Image Classification
He, Hangzhou (Peking University), Lu, Yanye (Peking University)
ClassificationDomain AdaptationExplainability and InterpretabilityConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A training-independent method for improving performance during testing is proposed, which enhances the performance of concept bottleneck models in new domains by identifying confusing concepts and applying masking and amplification.
TransSino: Prior Sinogram Pattern-Based Transformer for Limited-Angle CT Image Segmentation
Yoon, Jae Hyun (Chonnam National University), Yoo, Seok Bong (Chonnam National University)
SegmentationTransformerDiffusion modelImageBiomedical DataComputed Tomography
🎯 What it does: A Transformer-based LACT sinogram domain segmentation model called TransSino is proposed, which can achieve high-quality lesion segmentation in the absence of missing projection angles.
Treat: A Unified Text-guided Conditioned Deep Learning Model for Generalized Radiotherapy Treatment Planning
Kim, Sangwook (University Health Network), McIntosh, Chris (University Health Network)
TransformerBiomedical Data
🎯 What it does: A unified text-guided deep learning model, Treat, is proposed to predict dose distribution under various radiotherapy protocols, supporting personalized automatic radiotherapy planning.
Tree-based Semantic Losses: Application to Sparsely-supervised Large Multi-class Hyperspectral Segmentation
Wang, Junwen (King's College London), Vercauteren, Tom (King's College London)
SegmentationAnomaly DetectionConvolutional Neural NetworkImage
🎯 What it does: Two types of semantic loss functions based on label trees are proposed for sparse annotation multi-class hyperspectral image segmentation, and they are integrated with an OOD detection framework.
TReND: Transformer derived features and Regularized NMF for neonatal functional network Delineation
Mohapatra, Sovesh (Children's Hospital of Philadelphia), Huang, Hao (Beijing Normal University)
TransformerAuto EncoderBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a self-supervised Transformer-AE framework named TReND for extracting spatiotemporal features from neonatal rs-fMRI and achieving functional network partitioning through regularized non-negative matrix factorization combined with KMeans.
Trexplorer Super: Topologically Correct Centerline Tree Tracking of Tubular Objects in CT Volumes
Naeem, Roman (Chalmers University of Technology), Kahl, Fredrik (Chalmers University of Technology)
Object TrackingTransformerImageBiomedical DataComputed Tomography
🎯 What it does: This paper presents Trexplorer Super, an improved 3D CT tree-like tubular structure centerline tracking model based on DETR.
TRRG: Towards Truthful Radiology Report Generation With Cross-modal Disease Clue Enhanced Large Language Models
Wang, Yuhao (School of Computing and Information Technology Great Bay University), Yu, Zitong (School of Computing and Information Technology Great Bay University)
GenerationTransformerLarge Language ModelContrastive LearningImageTextMultimodalityElectronic Health Records
🎯 What it does: This paper proposes the TRRG framework, which utilizes cross-modal disease cues to enhance large language models for more realistic chest X-ray report generation.
Trustworthy Few-Shot Transfer of Medical VLMs through Split Conformal Prediction
Silva-Rodríguez, Julio (École de technologie supérieure Montréal), Dolz, Jose (CHU Sainte-Justine, University of Montreal)
ClassificationDomain AdaptationTransformerVision Language ModelContrastive LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper introduces Segmented Conformal Prediction (SCP) into few-shot medical visual language models and proposes Unsupervised Transductive Segmented Conformal Adaptation (SCA-T) to maintain coverage guarantees and improve the efficiency of the prediction set.
Tumor Microenvironment-Guided Fine-Tuning of Pathology Foundation Models for Esophageal Squamous Cell Carcinoma Immunotherapy Response Prediction
Lin, Yixuan (Xiamen University), Wang, Liansheng (Shanghai Changhai Hospital)
ClassificationSegmentationConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: A fine-tuning framework based on the tumor microenvironment (TME) for a pathological foundation model (PFM) has been designed and implemented to predict the response of esophageal squamous cell carcinoma (ESCC) patients to immunotherapy.
Tumor Segmentation with Heterogeneity Clustering in Non-contrast Breast MRI
Xie, Xinyu (Macao Polytechnic University), Shen, Dinggang (Shanghai United Imaging Intelligence Co., Ltd.)
SegmentationConvolutional Neural NetworkAuto EncoderImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a method for extracting tumor heterogeneity using DCE-MRI time intensity curve clustering, and achieves tumor segmentation on non-contrast breast MRI through a predictive model and a fusion segmentation network.
Two-Stage Generative Model for Intracranial Aneurysm Meshes with Morphological Marker Conditioning
Ding, Wenhao (Imperial College London), Yap, Choon Hwai (Imperial College London)
GenerationData SynthesisAuto EncoderMeshBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A two-stage conditional generative model, AneuG, was designed and implemented to synthesize brain aneurysm meshes and their parent vessels that meet specific morphological parameters.
U-Net Transplant: The Role of Pre-training for Model Merging in 3D Medical Segmentation
Lumetti, Luca (University of Modena and Reggio Emilia), Bolelli, Federico (University of Modena and Reggio Emilia)
SegmentationConvolutional Neural NetworkSupervised Fine-TuningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A task vector-based 3D medical segmentation model merging framework is proposed, and the impact of pre-training strategies on merging performance is analyzed in depth.
U-RWKV: Lightweight medical image segmentation with direction-adaptive RWKV
Ye, Hongbo (University of Science and Technology of China), Zhou, S. Kevin (Hohai University)
SegmentationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: Proposes the U-RWKV lightweight medical image segmentation framework, utilizing the RWKV structure and adaptive modules to achieve efficient long-range dependency modeling.
UDPET: Ultra-low Dose PET Imaging Challenge Dataset
Xue, Song (University of Bern), Shi, Kuangyu (University of Bern)
RestorationImageBiomedical DataPositron Emission TomographyBenchmark
🎯 What it does: This study designed and released the Ultra-Low Dose PET Imaging Challenge Dataset (UDPET) and its evaluation framework, aiming to promote the development and standardized assessment of low-dose whole-body PET image quality restoration algorithms.
UFO-3: Unsupervised Three-Compartment Learning for Fiber Orientation Distribution Function Estimation
Gao, Xueqing (Fudan University), Qiao, Yuchuan (Fudan University)
Convolutional Neural NetworkBiomedical DataMagnetic Resonance ImagingFibre Orientation Distribution
🎯 What it does: This paper presents UFO-3, an unsupervised three-compartment model combined with a U-Net framework for estimating the fiber orientation distribution function (fODF), which can achieve high-precision fODF estimation using only single-shell data.
Ultra-Low-Field MRI Enhancement via INR-Based Style Transfer
Islam, Kh Tohidul (Monash University), Chen, Zhaolin (Monash University)
Image TranslationRestorationNeural Radiance FieldImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an unsupervised framework that combines implicit neural representations (INR) with neural style transfer (NST) to transfer structural details from high-field MRI to ultra-low-field MRI, enhancing image quality.
UltraAD: Fine-Grained Ultrasound Anomaly Classification via Few-Shot CLIP Adaptation
Zhou, Yue (Computer Aided Medical Procedures (CAMP), TU Munich), Jiang, Zhongliang (Computer Aided Medical Procedures (CAMP), TU Munich)
ClassificationAnomaly DetectionTransformerPrompt EngineeringContrastive LearningImageBiomedical DataUltrasound
🎯 What it does: Adapt the CLIP model using a small number of ultrasound samples to jointly perform anomaly localization and fine-grained classification;
UltraRay: Introducing Full-Path Ray Tracing in Physics-Based Ultrasound Simulation
Duelmer, Felix (Technical University of Munich), Navab, Nassir (Technische Universität München)
MeshComputed TomographyUltrasound
🎯 What it does: A full-path ray tracing-based ultrasound physical simulation framework, UltraRay, is proposed, which can generate realistic B-mode images in real-time.
UltraTwin: Towards Cardiac Anatomical Twin Generation from Multi-view 2D Ultrasound
Yu, Junxuan (Shenzhen University), Yang, Xin (Shenzhen University)
SegmentationGenerationData SynthesisTransformerDiffusion modelAuto EncoderImageBiomedical DataComputed TomographyUltrasound
🎯 What it does: Using sparse multi-view 2D ultrasound images, a 3D cardiac anatomical digital twin is constructed through a diffusion transformer and an implicit autoencoder.
UltrON: Ultrasound Occupancy Networks
Wysocki, Magdalena (Technical University of Munich), Azampour, Mohammad Farid (Technical University of Munich)
Neural Radiance FieldImageMeshUltrasound
🎯 What it does: Utilizing the acoustic features of multi-view ultrasound B-mode images, a occupancy network is constructed to achieve 3D anatomical structure reconstruction.
UM-SAM: Unsupervised Medical Image Segmentation using Knowledge Distillation from Segment Anything Model
Fu, Jia (University of Electronic Science and Technology of China), Wang, Guotai (University Of Electronic Science And Technology Of China)
SegmentationKnowledge DistillationConvolutional Neural NetworkContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an unsupervised medical image segmentation framework called UM-SAM based on the Segment Anything Model (SAM), which generates pseudo-labels using SAM's everything mode, filters target pseudo-labels through shape priors and ROI feature clustering, and then trains a lightweight segmentation network using triple knowledge distillation.
Uncertainty-aware Diffusion and Reinforcement Learning for Joint Plane Localization and Anomaly Diagnosis in 3D Ultrasound
Huang, Yuhao (Shenzhen University), Ni, Dong (People's Hospital of Guangxi Zhuang Autonomous Region)
Anomaly DetectionOptimizationRecurrent Neural NetworkReinforcement LearningDiffusion modelImageBiomedical DataUltrasound
🎯 What it does: This paper proposes an intelligent system that integrates standard plane localization with the diagnosis of congenital uterine anomalies (CUA), capable of automatically locating the coronal plane in three-dimensional ultrasound images and providing diagnostic results.
Uncertainty-Aware Multi-Expert Knowledge Distillation for Imbalanced Disease Grading
Tong, Shuo (Zhejiang University), Wu, Jian (Zhejiang University)
Domain AdaptationKnowledge DistillationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A framework called 'Uncertainty-Aware Multi-Expert Knowledge Distillation (UMKD)' is proposed to transfer knowledge from multiple expert models to a single student model, addressing the issues of class imbalance and domain transfer in medical image classification.
Uncertainty-Aware Multimodal MRI Fusion for HIV-Associated Asymptomatic Neurocognitive Impairment Prediction
Chen, Zige (Nanjing University), Shan, Caifeng (Nanjing University)
ClassificationConvolutional Neural NetworkGraph Neural NetworkTransformerMultimodalityBiomedical DataMagnetic Resonance ImagingDiffusion Tensor Imaging
🎯 What it does: This paper proposes a tri-modal MRI fusion framework based on uncertainty perception (UMMF) for predicting HIV-related asymptomatic neurocognitive impairment (ANI).
Uncertainty-Supervised Interpretable and Robust Evidential Segmentation
Li, Yuzhu (Fudan University), Zhuang, Xiahai (Fudan University)
SegmentationExplainability and InterpretabilityImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: By incorporating self-supervised uncertainty loss based on image gradients and noise into the evidence deep learning segmentation model, the interpretability and robustness of uncertainty are enhanced.
UniCross: Balanced Multimodal Learning for Alzheimer’s Disease Diagnosis by Uni-modal Separation and Metadata-guided Cross-modal Interaction
Yin, Lisong (Beijing Institute of Technology), Yan, Tianyi (Beijing Institute of Technology)
ClassificationTransformerContrastive LearningMultimodalityBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease
🎯 What it does: Designed and evaluated a balanced multimodal learning framework called UniCross for Alzheimer's disease diagnosis and MCI conversion prediction.
UniMRG: Refining Medical Semantic Understanding Across Modalities via LLM-Orchestrated Synergistic Evolution
Xu, Hongyan (University of New South Wales), Wang, Dadong (University of New South Wales)
GenerationOptimizationNeural Architecture SearchTransformerLarge Language ModelVision Language ModelImageTextMultimodality
🎯 What it does: The UniMRG framework is proposed, utilizing LLM-guided multimodal augmentation (UMA) and a Medical Content Learner (MCL) to enhance the semantic understanding of cross-modal medical report generation.
UniOCTSeg: Towards Universal OCT Retinal Layer Segmentation via Hierarchical Prompting and Progressive Consistency Learning
Zhong, Jian, Tang, Xiaoying (Chinese University of Hong Kong)
SegmentationConvolutional Neural NetworkTransformerPrompt EngineeringBiomedical Data
🎯 What it does: This paper proposes a universal OCT retinal layer segmentation framework called UniOCTSeg, which can uniformly perform segmentation tasks on multi-granularity labeled datasets.
UniSegDiff: Boosting Unified Lesion Segmentation via a Staged Diffusion Model
Hu, Yilong (Dalian University of Technology), Lu, Huchuan (Dalian University of Technology)
SegmentationConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A unified lesion segmentation framework called UniSegDiff is proposed, which employs a staged diffusion model for training and inference.
Unisyn: A Generative Foundation Model for Universal Medical Image Synthesis across MRI, CT and PET
Wang, Yulin (ShanghaiTech University), Shen, Dinggang (Shanghai United Imaging Intelligence Co., Ltd.)
GenerationData SynthesisContrastive LearningImageBiomedical DataMagnetic Resonance ImagingComputed TomographyPositron Emission Tomography
🎯 What it does: The UniSyn framework is proposed, which can generate target medical images (MRI/CT/PET) with specified parameters based on any existing images combined with textual metadata (scanning parameters, clinical information).
Unleashing SAM for Few-Shot Medical Image Segmentation with Dual-Encoder and Automated Prompting
Pham, Cuong M. (Noi That University), Nguyen, Binh P. (Vietnam National University)
SegmentationImageMagnetic Resonance Imaging
🎯 What it does: This paper proposes an image segmentation method based on the Mean Shift algorithm, utilizing joint kernel density estimation of color and spatial features to iteratively update image pixels and ultimately obtain segmentation results.
Unleashing the Power of LLMs for Medical Video Answer Localization
Xiao, Junbin (National University of Singapore), Yao, Angela (National University of Singapore)
RecognitionRetrievalTransformerLarge Language ModelVision Language ModelVideoTextMultimodality
🎯 What it does: A zero-shot medical teaching video answer localization method is proposed, utilizing a combination of large language models and multimodal large language models to perform temporal answer localization using subtitles and visual descriptions.
Unleashing Vision Foundation Models for Coronary Artery Segmentation: Parallel ViT-CNN Encoding and Variational Fusion
Dong, Caixia, Xu, Songhua
SegmentationConvolutional Neural NetworkTransformerImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a parallel encoding architecture based on visual foundation models for coronary artery segmentation.
Unpaired Multi-Site Brain MRI Harmonization with Image Style-Guided Latent Diffusion
Wu, Mengqi (University of North Carolina at Chapel Hill), Liu, Mingxia (University of North Carolina at Chapel Hill)
Image HarmonizationSegmentationDiffusion modelAuto EncoderImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A unpaired multi-site brain MRI image harmonization framework called UMH is proposed, which employs a two-stage conditional latent diffusion model and CLIP-style guided refinement to achieve seamless conversion between a unified domain and target site styles.
Unraveling Brainstem Deformations in Joubert Syndrome: A Statistical Shape Analysis of MRI-Derived Structures
Maccarone, Francesca (University of Milano-Bicocca), Melzi, Simone (University of Milano-Bicocca)
ClassificationSegmentationImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A pipeline for brainstem shape processing based on Statistical Shape Analysis (SSA) was developed to detect Joubert syndrome using MRI surface data.
Unsupervised Anomaly Detection on Preclinical Liver H&E Whole Slide Images using Graph based Feature Distillation
Li, Lin (Merck & Co., Inc.), Chen, Antong (Merck & Co., Inc.)
Anomaly DetectionKnowledge DistillationTransformerImage
🎯 What it does: This study proposes the GraphTox model, which achieves unsupervised anomaly detection of preclinical liver H&E whole slide images through graphical feature distillation.
Unsupervised Cardiac Video Translation Via Motion Feature Guided Diffusion Model
Deb, Swakshar (University of Virginia), Zhang, Miaomiao (University of Virginia)
Image TranslationSegmentationGenerationDiffusion modelVideoBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An unsupervised cardiac video translation model MFD-V2V has been developed, capable of converting low signal-to-noise ratio and low-contrast DENSE CMR sequences into high-quality, high-contrast cine CMR sequences.
Unsupervised Discovery of Spatiotypes and Context-Aware Graph Neural Networks for Modeling Clinical Endpoints
Dawood, Muhammad (University of Oxford), Rittscher, Jens (University of Oxford)
Graph Neural NetworkBiomedical Data
🎯 What it does: This paper proposes a Band Descriptor based on concentric rings to quantify the relative abundance of different cell types surrounding single cells. Utilizing this descriptor, substructures of the local cellular microenvironment (spatiotypes) were discovered in spatial transcriptomic samples of pulmonary fibrosis, and these were incorporated as node features into a graph neural network (GNN) to enhance the accuracy of clinical outcome predictions.
Unsupervised Learning-Based Susceptibility Artifact Correction for Diffusion-Weighted MRI in Multiple Organs
Qiu, Shihan (Siemens Healthineers), Nadar, Mariappan S. (Siemens Healthineers)
RestorationTransformerBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A method for EPI susceptibility artifact correction based on unsupervised learning is proposed, which estimates the displacement field through a pair of reversed phase-encoded (reversed-PE) images and performs distortion correction on multi-organ diffusion-weighted MRI (DWI).
Unsupervised OCT image interpolation using deformable registration and generative models
Wei, Shuwen (Johns Hopkins University), Carass, Aaron (Johns Hopkins University)
Image TranslationRestorationGenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An unsupervised OCT image interpolation method is proposed, combining deformation registration with conditional DDPM to generate structurally consistent and realistic interpolated slices.
Unsupervised Quality Control and Enhancement of Polyp Segmentation in Colonoscopy Videos using Spatiotemporal Consistency
Li, Yujia (Nanjing University of Science and Technology), Zhang, Yizhe (Nanjing University of Science and Technology)
SegmentationVideo
🎯 What it does: This paper proposes an unsupervised quality control and enhancement framework for polyp segmentation in colonoscopy videos, utilizing the video segmentation base model SAM2 to propagate segmentation results over time, calculate the spatiotemporal consistency score (SQA), and perform re-segmentation of low-quality frames based on high-quality frames.
Unsupervised Structure-Geometric Consistency for Monocular Endoscopic Depth Overestimation
Fan, Wenkang (Xiamen University), Luo, Xiongbiao (Xiamen University)
Depth EstimationConvolutional Neural NetworkTransformerOptical FlowVideo
🎯 What it does: An unsupervised structural geometric consistency framework is proposed, utilizing a small dense convolutional hierarchical transformer (DCHT) to simultaneously estimate monocular endoscopic depth and camera pose, addressing depth overestimation and weak texture issues through structure-aware photometric consistency and 3D geometric consistency loss.
UNSURF: Uncertainty Quantification for Cortical Surface Reconstruction of Clinical Brain MRIs
Gopinath, Karthik (A. A. Martinos Center for Biomedical Imaging Massachusetts General Hospital), Iglesias, Juan Eugenio (A. A. Martinos Center for Biomedical Imaging Massachusetts General Hospital)
ClassificationSegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A metric for cortical surface reconstruction uncertainty named UNSURF is proposed for quality control and downstream analysis of clinical MRI scans.
Untangling Vascular Trees for Surgery and Interventional Radiology
Houry, Guillaume (Inria, Université Paris Cité, Inserm, HeKA), Feydy, Jean (Inria, Université Paris Cité, Inserm, HeKA)
SegmentationOptimizationImageBiomedical DataComputed Tomography
🎯 What it does: This study proposes a method for rapidly generating two-dimensional planar embeddings of vascular trees for surgical planning and catheter navigation in interventional radiology.
UWT-Net: Mining low-frequency feature information for medical image segmentation
Zhang, Pengcheng (Sichuan Agricultural University), Peng, Ran (Southwestern University of Finance and Economics)
SegmentationConvolutional Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: The UWT-Net model is proposed, which utilizes the IMFIE block combined with Wavelet convolution to extract low-frequency features in medical images, achieving end-to-end segmentation.
V2T-CoT: From Vision to Text Chain-of-Thought for Medical Reasoning and Diagnosis
Wang, Yuan (Zhejiang University), Liu, Zuozhu (Zhejiang University)
Object DetectionExplainability and InterpretabilityTransformerLarge Language ModelVision Language ModelImageTextMultimodalityBiomedical DataChain-of-Thought
🎯 What it does: The V2T-CoT method is proposed, which applies Vision-to-Text Chain-of-Thought (CoT) to medical visual question answering (Med-VQA) by automatically locating disease-related areas and generating interpretable diagnostic reasoning paths based on this.
VAMPIRE: Uncovering Vessel Directional and Morphological Information from OCTA Images for Cardiovascular Disease Risk Factor Prediction
Wang, Lehan (Hong Kong University of Science and Technology), Li, Xiaomeng (Southern Medical University)
ClassificationSegmentationConvolutional Neural NetworkTransformerLarge Language ModelImageBiomedical Data
🎯 What it does: A multi-task framework called VAMPIRE based on OCTA images is proposed, which can simultaneously predict the 10-year risk of cardiovascular disease and four blood indicators (high blood sugar, high cholesterol, high triglycerides, high blood pressure);
VAP-Diffusion: Enriching Descriptions with MLLMs for Enhanced Medical Image Generation
Huang, Peng (Fudan University), Guo, Yi (Fudan University)
GenerationData SynthesisLarge Language ModelPrompt EngineeringDiffusion modelImageBiomedical DataChain-of-Thought
🎯 What it does: This paper proposes a framework for generating medical images by utilizing multimodal large language models to generate visual attribute prompts, combined with diffusion models;
Variational Visible Layers: A Practical Framework for Uncertainty Estimation
Abboud, Zeinab (Polytechnique Montreal), Kadoury, Samuel (CHUM Hospital Research Center)
ClassificationSegmentationConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: In medical image classification and segmentation tasks, a method is proposed to introduce Variational Visible Layers (VVL) at the first and/or last layer of the network, which allows for post-hoc fine-tuning on existing models or joint training, achieving lightweight uncertainty estimation.
Various Attention Mechanism Graph Convolutional Network with Multi-Source Domain Adaptation for Cross-Subject EEG Emotion Recognition
Shi, Shuo (Hebei University of Technology), Hao, Xiaoke (Hebei University of Technology)
RecognitionDomain AdaptationGraph Neural NetworkBiomedical Data
🎯 What it does: This paper proposes a model that combines various attention mechanisms with graph convolutional networks and multi-source domain adaptation for cross-subject emotion recognition.
Vascular Photoacoustic Volume Registration via 2D Feature Matching with Reverse Mapping Based on Maximum Intensity Projection
Liao, Junda (University of Tokyo), Sato, Imari (Keio University)
SegmentationOptimizationImageBiomedical DataUltrasound
🎯 What it does: An affine registration framework for vascular photoacoustic volumes is proposed, combining 2D feature matching + inverse mapping, feature-guided intensity resampling, and hair removal preprocessing to address issues such as sparsity, blurriness, network changes, and hair interference.
VBCD: A Voxel-Based Framework for Personalized Dental Crown Design
Wei, Linda (Chinese University of Hong Kong), Li, Hongsheng (Sensetime Research)
SegmentationGenerationConvolutional Neural NetworkPoint Cloud
🎯 What it does: This paper proposes a voxel-based coarse-to-fine two-stage framework (VBCD) that can automatically generate personalized dental crowns from intraoral scanning data.
Vector-Quantization-Driven Active Learning for Efficient Multi-Modal Medical Segmentation with Cross-Modal Assistance
Du, Xiaofei (Fudan University), Song, Zhijian (Fudan University)
SegmentationAuto EncoderMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A dual-modal segmentation and active learning framework based on vector quantization (VQ-BEGAL) is proposed, achieving feature decoupling and reliable uncertainty estimation.
Veriserum: A dual-plane fluoroscopic dataset with knee implant phantoms for deep learning in medical imaging
Wang, Jinhao (ETH Zürich), Taylor, William R. (ETH Zürich)
Pose EstimationImage
🎯 What it does: This paper constructs Veriserum, the first open-source dual-plane knee implant X-ray dataset, which contains approximately 110,000 images and corresponding pose annotations.