MICCAI 2025 Papers — Page 4
International Conference on Medical Image Computing and Computer-Assisted Intervention · 1027 papers
Enhancing Radiology Report Interpretation through Modality-Specific RadGraph Fine-Tuning
Guan, Haoyue (Johns Hopkins University), Bai, Harrison (Brown University)
TransformerSupervised Fine-TuningMultimodalityBiomedical DataMagnetic Resonance ImagingComputed TomographyUltrasound
🎯 What it does: A dataset of expert annotated reports for four imaging modalities (cardiac MRI, abdominal ultrasound, head CT, and CT pulmonary angiography) was constructed, and based on this, RadGraph was fine-tuned for modality-specific tasks;
Enhancing Soft Tissue Sarcoma Classification by Mitigating Patient-Specific Bias in Whole Slide Images
Lin, Weiping (Xiamen University), Wang, Liansheng (Shanghai Changhai Hospital)
ClassificationTransformerContrastive LearningImage
🎯 What it does: Utilizing a multi-instance learning framework combined with supervised contrastive learning and propensity score matching to eliminate patient-specific biases in WSI, thereby improving the classification performance of soft tissue sarcoma subtypes.
Enhancing WSI-Based Survival Analysis with Report-Auxiliary Self-Distillation
Wang, Zheng (Xiamen University), Wang, Liansheng (Southern Medical University)
Knowledge DistillationTransformerLarge Language ModelSupervised Fine-TuningImageTextBiomedical Data
🎯 What it does: This paper proposes a WSI survival analysis framework called Rasa, based on report-assisted self-distillation, which extracts precise report text using LLM to guide feature selection and data augmentation, thereby improving the survival prediction performance of WSI.
Enjoying Information Dividend: Gaze Track-based Medical Weakly Supervised Segmentation
Wang, Zhisong (Ningbo Institute of Northwestern Polytechnical University), Xia, Yong (Northwestern Polytechnical University)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The GradTrack framework is proposed, utilizing doctors' eye movement trajectory information (fixation points, dwell time, temporal order) to achieve weakly supervised medical image segmentation.
Equitable Federated Learning with NCA
Lemke, Nick (Technical University of Darmstadt), Mukhopadhyay, Anirban (Carl Zeiss AG)
SegmentationFederated LearningSafty and PrivacyComputational EfficiencyImageBiomedical DataUltrasound
🎯 What it does: This paper proposes FedNCA—a federated learning framework based on a lightweight neural cellular automaton (Med-NCA) that can perform medical image segmentation tasks on edge devices with low computing power and low bandwidth (such as mobile phones), while supporting homomorphic encryption to ensure secure aggregation on the server.
ESPNet: Edge-Aware Feature Shrinkage Pyramid for Polyp Segmentation
Toman, Raneem (University of Leeds), Ali, Sharib (University of Leeds)
SegmentationTransformerImage
🎯 What it does: This paper presents ESPNet, a Transformer-based edge-aware feature shrinkage pyramid network for polyp segmentation in multi-center and diverse populations.
Estimating Bone Mineral Density and Muscle Mass from EOS Low Dose X-ray Imaging System
Suehara, Kazuki (Nara Institute of Science and Technology), Sato, Yoshinobu (Nara Institute of Science and Technology)
SegmentationDomain AdaptationSupervised Fine-TuningGenerative Adversarial NetworkImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a deep learning framework that utilizes low-dose dual-plane X-rays combined with a small amount of CT data to estimate bone density and muscle mass.
EUReg: End-to-end Framework for Efficient 2D-3D Ultrasound Registration
Wang, Haiqiao (Shenzhen University), Wang, Yi (Shenzhen University)
OptimizationComputational EfficiencyConvolutional Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: We propose EUReg, an end-to-end real-time 2D-3D ultrasound registration framework that addresses the shortcomings of existing methods in terms of accuracy, efficiency, and overfitting.
Exemplar Med-DETR: Toward Generalized and Robust Lesion Detection in Mammogram Images and Beyond
Bhat, Sheethal (Friedrich-Alexander-Universität), Maier, Andreas (Friedrich-Alexander-Universität Erlangen)
Object DetectionTransformerContrastive LearningImageMultimodality
🎯 What it does: This paper proposes Exemplar Med-DETR, which utilizes class-specific exemplar features to achieve multimodal contrastive detection, significantly improving lesion localization performance in mammography, chest X-ray, and angiography images.
Explain Any Pathological Concept: Discovering Hierarchical Explanations for Pathology Foundation Models
Xu, Shuting (Nankai University), Chen, Hao (Harvard Medical School)
ClassificationSegmentationExplainability and InterpretabilityImageBiomedical Data
🎯 What it does: This paper proposes a hierarchical concept explanation framework (HCE) that combines expert-general collaborative segmentation, lightweight proxy models, and Shapley computation to automatically extract three layers of concepts (cell/unit/region) from pathological images, and visualizes the contributions of each layer through ShapMap to explain the decisions of the base models (UNI, CONCH, Virchow).
Explainable ADHD Diagnostic Framework Using Weakly-Supervised Action Recognition
Fan, Ninghan (Zhejiang University), Zhu, Qiang (Zhejiang University)
Anomaly DetectionExplainability and InterpretabilityVideoBiomedical Data
🎯 What it does: Proposes the EDWAR framework, which integrates executive function test indicators with weakly supervised action recognition to achieve interpretable ADHD diagnosis.
Explainable Classifier for Malignant Lymphoma Subtyping via Cell Graph and Image Fusion
Nishiyama, Daiki (Institute of Science Tokyo), Sakuma, Jun (Institute of Science Tokyo)
ClassificationExplainability and InterpretabilityConvolutional Neural NetworkGraph Neural NetworkMixture of ExpertsImageMultimodalityBiomedical Data
🎯 What it does: An interpretable multimodal MIL framework is proposed, utilizing cell graphs and image fusion to classify three subtypes of malignant lymphoma (DLBCL, FL, Reactive), and provides class-level ROI and explanations of cell frequency and spatial distribution.
Explainable Integrative Bipartite Graph Convolutional Neural Network for Predicting Ejection Fraction in Echocardiography
Lee, Seungeun (Klleon), Kang, Mingon (University of Nevada Las Vegas)
Explainability and InterpretabilityConvolutional Neural NetworkGraph Neural NetworkAuto EncoderVideoMultimodalityTabularUltrasound
🎯 What it does: An interpretable bilateral graph convolutional neural network, IBi-GNN, is proposed to integrate cardiac ultrasound videos and demographic features (age, gender, BMI) to predict left ventricular ejection fraction.
Exploring Text-enhanced Mixture-of-Experts for Semi-supervised Medical Image Segmentation with Composite Data
Zeng, Qingjie (Northwestern Polytechnical University), Xia, Yong (Northwestern Polytechnical University)
SegmentationConvolutional Neural NetworkMixture of ExpertsImageTextMultimodalityBiomedical DataComputed Tomography
🎯 What it does: This paper studies a text-enhanced mixture of experts model, TextMoE, for semi-supervised medical image segmentation of combined CT and X-ray images.
Exploring the Design Space of 3D MLLMs for CT Report Generation
Baharoon, Mohammed (Vector Institute for Artificial Intelligence), Wang, Bo (Vector Institute for Artificial Intelligence)
GenerationTransformerLarge Language ModelSupervised Fine-TuningMultimodalityBiomedical DataComputed Tomography
🎯 What it does: This paper systematically evaluates the design space of 3D multimodal large language models (MLLM) in CT report generation and proposes two knowledge enhancement methods to improve the completeness and quality of reports.
Exploring the feasibility of zero-shot super-resolution in preclinical imaging
Gharib, Omar A. M. (Johns Hopkins University), Carass, Aaron (Johns Hopkins University)
RestorationSuper ResolutionDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The study achieves zero-shot super-resolution of small animal MRI without training data, proposing the Biplanar DDNM Averaging (BiDA) method.
Exposing and Mitigating Calibration Biases and Demographic Unfairness in MLLM Few-Shot In-Context Learning for Medical Image Classification
Shen, Xing (McGill University), Arbel, Tal (McGill University)
ClassificationTransformerLarge Language ModelImageMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study investigates the calibration bias and fairness issues of multimodal large language models in few-shot contextual learning for medical imaging, and proposes a training-free second-order calibration method called CALIN.
F2PASeg: Feature Fusion for Pituitary Anatomy Segmentation in Endoscopic Surgery
Chen, Lumin (Centre for Artificial Intelligence and Robotics, Hong Kong Institute of Science Innovation, Chinese Academy of Sciences), Liu, Hongbin (Centre for Artificial Intelligence and Robotics, Hong Kong Institute of Science Innovation, Chinese Academy of Sciences)
SegmentationTransformerImageBiomedical Data
🎯 What it does: Proposes F2PASeg, which combines a Feature Fusion module for real-time anatomical structure semantic segmentation in endoscopic pituitary surgery.
Facial Appearance Prediction with Conditional Multi-scale Autoregressive Modeling for Orthognathic Surgical Planning
Lee, Jungwook (Rensselaer Polytechnic Institute), Yan, Pingkun (Rensselaer Polytechnic Institute)
GenerationImageComputed Tomography
🎯 What it does: The CAMOS framework is proposed, which uses a conditional autoregressive multiscale model to directly predict the optimal 3D appearance post-surgery from the preoperative facial deformity of patients.
Fair-MoE: Medical Fairness-Oriented Mixture of Experts in Vision-Language Models
Wang, Peiran (Zhejiang University), Liu, Zuozhu (Zhejiang University)
TransformerMixture of ExpertsVision Language ModelContrastive LearningImageTextMultimodality
🎯 What it does: The Fair-MoE framework is proposed, which enhances fairness and diagnostic effectiveness in medical vision-language models through two modules: FO-MoE and FOL.
Fairness-Aware vCDR-Controlled Generation for Glaucoma Diagnosis
Wang, Ziheng (University of Exeter), Meng, Yanda (University Of Exeter)
SegmentationGenerationData SynthesisDiffusion modelImage
🎯 What it does: The GlaucoDiff model is proposed, which can achieve bidirectional generation of healthy and glaucoma retinal images by controlling the vertical cup-to-disc ratio (vCDR), thereby enriching the dataset and enhancing diagnostic fairness.
Faster, Self-Supervised Super-Resolution for Anisotropic Multi-View MRI Using a Sparse Coordinate Loss
Schlereth, Maja (Friedrich-Alexander-Universität Erlangen-Nürnberg), Breininger, Katharina (Friedrich-Alexander-Universität Erlangen-Nürnberg)
RestorationSuper ResolutionConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A self-supervised multi-view super-resolution network called tripleSR is proposed, which fuses two orthogonal low-resolution MRI images using sparse coordinate loss to generate a unified high-resolution image.
FDAS: Foundation Model Distillation and Anatomic Structure-aware Multi-task Learning for Self-Supervised Medical Image Segmentation
Qi, Xiaoran (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: A self-supervised medical image segmentation framework FDAS based on foundational model distillation and anatomy structure-aware multi-task learning is proposed.
FDF-VQVAE: A Frequency Disentanglement and Fusion Learning Framework for Multi-Sequence MRI Enhancement
Xie, Xinghe (Macao Polytechnic University), Tan, Tao (Macao Polytechnic University)
RestorationSuper ResolutionConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an interpretable multi-sequence MRI enhancement framework FDF-VQVAE, which achieves image denoising, super-resolution, and motion artifact removal through frequency domain feature separation and fusion.
FEAT: Full-Dimensional Efficient Attention Transformer for Medical Video Generation
Wang, Huihan (Beihang University), Xu, Yan (Beihang University)
GenerationData SynthesisTransformerDiffusion modelVideoBiomedical Data
🎯 What it does: This paper proposes a Full-Dimensional Efficient Attention Transformer (FEAT) for medical video generation, capable of modeling global dependencies across spatial, temporal, and channel dimensions simultaneously.
Feature Copy-Paste Network for Lung Cancer EGFR Mutation Status Prediction in CT images
Huang, Xingyu (Beihang University), Tian, Jie (Beihang University)
ClassificationConvolutional Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: A deep learning model called FCPNet is proposed to predict the EGFR mutation status of lung cancer patients using CT images.
Feature Mixing Approach for Detecting Intraoperative Adverse Events in Laparoscopic Roux-en-Y Gastric Bypass Surgery
Bose, Rupak (University of Strasbourg), Padoy, Nicolas (University of Strasbourg, CNRS, INSERM, ICube, UMR7357)
Anomaly DetectionTransformerGenerative Adversarial NetworkVideo
🎯 What it does: A deep learning framework named BetaMixer is proposed for real-time detection and quantitative assessment of intraoperative adverse events (bleeding, mechanical injury, and thermal injury) during laparoscopic Roux-Y gastric bypass surgery.
FedAMM: Federated Learning for Brain Tumor Segmentation with Arbitrary Missing Modalities
Shi, Yukun (Hangzhou Dianzi University), Zhou, Ye (Westlake University)
SegmentationFederated LearningKnowledge DistillationConvolutional Neural NetworkMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study investigates brain tumor segmentation in a federated learning environment and proposes the FedAMM framework to address the issue of arbitrary missing modalities.
FedCLAM: Client Adaptive Momentum with Foreground Intensity Matching for Federated Medical Image Segmentation
Siomos, Vasilis (CitAI Research Centre), Tarroni, Giacomo (CitAI Research Centre)
SegmentationFederated LearningConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: FedCLAM is proposed under the federated learning framework to address the heterogeneity issue in medical image segmentation.
FedWSIDD: Federated Whole Slide Image Classification via Dataset Distillation
Jin, Haolong (Huazhong University of Science and Technology), Hu, Yingzi (Huazhong University of Science and Technology)
ClassificationFederated LearningKnowledge DistillationConvolutional Neural NetworkImage
🎯 What it does: This paper proposes FedWSIDD, which achieves personalized training for WSI classification by transmitting synthetic Whole Slide Images instead of model parameters in federated learning.
Feeling the Stakes: Realism and Ecological Validity in User Research for Computer-Assisted Interventions
Cho, Sue Min (Johns Hopkins University), Unberath, Mathias (University of Arkansas)
Image
🎯 What it does: Compared the user behavior and subjective experience of completing 2D/3D registration assessment tasks in a regular office laboratory versus a high-fidelity simulated operating room (MockOR).
Fetuses Made Simple: Modeling and Tracking of Fetal Shape and Pose
Liu, Yingcheng (Massachusetts Institute of Technology), Golland, Polina (Boston Children's Hospital)
SegmentationPose EstimationOptimizationImageTime SeriesBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A 3D fetal shape and posture statistical model based on SMPL has been constructed, achieving precise alignment and visualization of fetal 3D shapes through alternating optimization of posture and shape on MRI time series; an automated method for conventional fetal measurements is also provided.
Few-Shot, Now for Real: Medical VLMs Adaptation without Balanced Sets or Validation
Silva-Rodríguez, Julio (École de technologie supérieure Montréal), Ben Ayed, Ismail (ÉTS Montréal)
ClassificationSegmentationDomain AdaptationTransformerVision Language ModelContrastive LearningMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a new unvalidated, unbalanced support set evaluation framework and a training-free linear probe SS-Text+ for few-shot adaptation in medical visual language models, aimed at achieving robust few-shot adaptation in real medical scenarios.
FilterDiff: Noise-free Frequency-domain Diffusion Models for Accelerated MRI Reconstruction
Song, Tao (Fudan University), Zhang, Shaoting (Zhongda Hospital)
RestorationTransformerDiffusion modelBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A noise-free frequency domain diffusion model called FilterDiff is proposed to accelerate MRI reconstruction.
FIND-Net – Fourier-Integrated Network with Dictionary Kernels for Metal Artifact Reduction
Tasharofi, Farid (Friedrich-Alexander-Universität Erlangen), Maier, Andreas (Friedrich-Alexander-Universität Erlangen)
RestorationOptimizationConvolutional Neural NetworkImageComputed Tomography
🎯 What it does: The FIND-Net framework is proposed, which effectively reduces CT metal artifacts by integrating frequency domain and spatial domain convolutions.
Fine-Grained Rib Fracture Diagnosis with Hyperbolic Embeddings: A Detailed Annotation Framework and Multi-Label Classification Model
Pate, Shripad (Indian Institute of Technology Jodhpur), Mishra, Deepak (Government Medical College Srinagar)
ClassificationObject DetectionConvolutional Neural NetworkContrastive LearningImageTextComputed Tomography
🎯 What it does: This paper constructs a fine-grained rib fracture diagnosis framework, first using a Faster R-CNN-based detector to locate fractures, and then employing a multi-head classifier and hyperplane multimodal embedding to achieve multi-label classification across four dimensions: location, displacement, morphology, and quantity.
Fine-tuning Vision Language Models with Graph-based Knowledge for Explainable Medical Image Analysis
Li, Chenjun (Cornell University), Paetzold, Johannes C. (Cornell Tech)
ClassificationExplainability and InterpretabilityGraph Neural NetworkSupervised Fine-TuningVision Language ModelImageBiomedical Data
🎯 What it does: This paper proposes a method that combines a heterogeneous graph of biological information with a visual language model, utilizing graph neural networks to stage diabetic retinopathy on OCTA images and generate interpretable diagnostic reports.
Finer Disentanglement of Aleatoric Uncertainty Can Accelerate Chemical Histopathology Imaging
Oh, Ji-Hun (University of Illinois Urbana Champaign), Bhargava, Rohit (University of Illinois Urbana Champaign)
SegmentationOptimizationConvolutional Neural NetworkSupervised Fine-TuningBiomedical Data
🎯 What it does: An adaptive chemical imaging strategy is proposed: first, low-information scanning samples are used to identify high-variance uncertainty regions, and then high-quality resampling is applied only in these regions to enhance segmentation performance.
Fit Pixels, Get Labels: Meta-Learned Implicit Networks for Image Segmentation
Vyas, Kushal, Balakrishnan, Guha (Rice University)
SegmentationMeta LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study presents MetaSeg, an implicit neural network framework for medical image segmentation that utilizes meta-learning.
Flexibly Distilled 3D Rectified Flow with Anatomical Constraints for Developmental Infant Brain MRI Prediction
Wang, Haifeng (Xian Jiaotong University), Ma, Jianhua (Pazhou Lab)
SegmentationGenerationRectified FlowImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Proposes the Flexibly Distilled 3D Rectified Flow (FDRF) framework for predicting brain images and tissue segmentation at 12 or 24 months from 6 months of brain MRI;
Flip Distribution Alignment VAE for Multi-Phase MRI Synthesis
Kui, Xiaoyan (Central South University), Zou, Beiji (Central South University)
GenerationData SynthesisAuto EncoderGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A lightweight Flip Distribution Alignment Variational Autoencoder (FDA-VAE) has been designed and implemented to achieve multi-phase CE MRI image synthesis.
Flow Matching for Medical Image Synthesis: Bridging the Gap Between Speed and Quality
Yazdani, Milad (University of British Columbia), Shahriari, Dena (University of British Columbia)
SegmentationGenerationData SynthesisConvolutional Neural NetworkDiffusion modelFlow-based ModelImageBiomedical DataMagnetic Resonance ImagingUltrasound
🎯 What it does: A medical image synthesis framework called MOTFM based on optimal transport flow matching is proposed, which can quickly generate high-quality medical images under various modalities, dimensions, and conditions, and can be used for tasks such as generation, segmentation, and denoising.
FluoroSAM: A Language-promptable Foundation Model for Flexible X-ray Image Segmentation
Killeen, Benjamin D. (Johns Hopkins University), Unberath, Mathias (University of Arkansas)
SegmentationTransformerPrompt EngineeringImageComputed Tomography
🎯 What it does: This paper presents FluoroSAM, a language prompt-based foundational model capable of segmenting X-ray images through natural language prompts, achieving segmentation of various organs and tools through training from scratch.
FMM-Diff: A Feature Mapping and Merging Diffusion Model for MRI Generation with Missing Modality
Zhong, Wenjin (Macquarie University), Liu, Sidong (Macquarie University)
GenerationData SynthesisDiffusion modelMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A framework called FMM-Diff based on diffusion models is proposed, which can generate high-level sequences (such as DWI or T1ce) through feature mapping and fusion modules when multi-modal MRI is missing.
Focus on Texture: Rethinking Pre-training in Masked Autoencoders for Medical Image Classification
Madan, Chetan (Indian Institute of Technology Delhi), Arora, Chetan (PGIMER Chandigarh)
ClassificationTransformerAuto EncoderImageBiomedical DataComputed TomographyUltrasound
🎯 What it does: This paper proposes a GLCM-based Masked Autoencoder (GLCM-MAE) pre-training framework that utilizes texture information to enhance the performance of medical image classification models.
FOCUS: Feature Replay with Optimized Channel-Consistent Dropout for U-Net Skip-Connections
Joham, Simon Johannes (Medical University of Graz), Urschler, Martin (Medical University of Graz)
SegmentationDomain AdaptationSafty and PrivacyConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a feature replay framework named FOCUS for domain incremental continuous medical image segmentation, which retains the skip connections of U-Net while meeting privacy and storage constraints.
Forget-MI: Machine Unlearning for Forgetting Multimodal Information in Healthcare Settings
Hardan, Shahad (Mohamed bin Zayed University of Artificial Intelligence), Yaqub, Mohammad (Mohamed bin Zayed University of Artificial Intelligence)
ClassificationData-Centric LearningConvolutional Neural NetworkMultimodalityBiomedical DataElectronic Health Records
🎯 What it does: The Forget-MI method is proposed, achieving machine unlearning of multimodal medical data while successfully forgetting patient data without compromising the overall performance of the model.
Foundation-Model-Boosted Multimodal Learning for fMRI-based Neuropathic Pain Drug Response Prediction
Fan, Wenrui (University of Sheffield), Zhou, Shuo (University of Sheffield)
Drug DiscoveryConvolutional Neural NetworkTransformerMultimodalityTime SeriesBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A multimodal learning framework FMM TC based on foundational models is proposed and implemented to predict drug responses in patients with neuropathic pain.
FoundBioNet: A Foundation-Based Model for IDH Genotyping of Glioma from Multi-Parametric MRI
Farahani, Somayeh (Tehran University of Medical Sciences), Liu, Sidong (Macquarie University)
ClassificationSegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper presents FoundBioNet, a foundational model based on SWIN-UNETR, which non-invasively predicts IDH mutation status in multiparametric MRI by integrating tumor-aware feature encoding and T2-FLAIR differences.
FPN-in-FPN: A Nested Multi-Scale Aggregation Network for Polyp Segmentation
Ye, Jin (Monash University), Cai, Jianfei (Monash University)
SegmentationConvolutional Neural NetworkImage
🎯 What it does: A nested multi-scale aggregation network (FPN-in-FPN) is designed and implemented for colon polyp segmentation, combining bidirectional feature fusion and deep supervision.
Frequency Strikes Back: Boosting Parameter-Efficient Foundation Model Adaptation for Medical Imaging
Ly, Son T. (University of Houston), Nguyen, Hien V. (University of Houston)
ClassificationDomain AdaptationTransformerSupervised Fine-TuningImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes a frequency domain fine-tuning module called FreqFiT, which is inserted between ViT blocks to enhance the performance of parameter-efficient fine-tuning.
Frequency-domain Multi-modal Fusion for Language-guided Medical Image Segmentation
Yu, Bo (Anhui University), Wang, Liang (Capital Medical University Affiliated Beijing Friendship Hospital)
SegmentationConvolutional Neural NetworkImageTextMultimodalityBiomedical DataComputed Tomography
🎯 What it does: This paper proposes FMISeg, a language-guided medical image segmentation model that achieves fine segmentation through bidirectional interaction of high and low-frequency visual features and text features.
Frequency-enhanced Multi-granularity Context Network for Efficient Vertebrae Segmentation
Shi, Jian (Dalian University of Technology), Li, Haojie (Shandong University of Science and Technology)
SegmentationImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A multi-granularity context network based on frequency domain enhancement, FMC-Net, is proposed for efficiently and accurately segmenting individual vertebrae in 3D CT and MRI images.
From Generalist to Specialist: Distilling a Mixture of Foundation Models for Domain-specific Medical Image Segmentation
Li, Qing (Fudan University), Wang, Chengyan (Fudan University)
SegmentationKnowledge DistillationImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: Knowledge from multiple general medical image segmentation base models is compressed into a specialized small model through multi-source knowledge distillation, without the need for annotations.
From Pixels to Prognosis: A Multi-Modal Attention-based Framework for Visceral Adipose Tissue Estimation
Maqsood, Arooba (Edith Cowan University), Gilani, Syed Zulqarnain (University of Western Australia)
TransformerImageMultimodalityBiomedical DataComputed Tomography
🎯 What it does: A multimodal attention framework was constructed to predict visceral adipose tissue (VAT) quality using lateral DXA scan images along with demographic information such as patient age, weight, and height.
From Sight to Skill: A Surgeon-Centered Augmented Reality System for Ureteroscopy Training
Atoum, Jumanh (Vanderbilt University), Wu, Jie Ying (Florida International University)
Biomedical Data
🎯 What it does: This study developed and evaluated a renal pelvis training system based on AR eye tracking, providing real-time visual guidance to trainees using three custom markers.
From Slices to Volumes: Multi-Scale Fusion of 2D and 3D Features for CT Scan Report Generation
Hosseini, Abdullah (Weill Cornell Medicine-Qatar), Serag, Ahmed (Weill Cornell Medicine-Qatar)
GenerationTransformerLarge Language ModelContrastive LearningImageTextMultimodalityComputed Tomography
🎯 What it does: A Slice-Attentive Multi-Modal Fusion (SAMF) framework is proposed, which combines a pre-trained 2D self-supervised encoder and a 3D aggregator to perform multi-scale feature fusion on CT scans, generating more accurate lung CT reports and completing multi-choice question-answering tasks.
From Variability To Accuracy: Conditional Bernoulli Diffusion Models with Consensus-Driven Correction for Thin Structure Segmentation
An, Jinseo (Seoul Women's University), Hong, Helen (Seoul Women's University)
SegmentationDiffusion modelImageBiomedical DataComputed Tomography
🎯 What it does: Combining the conditional Bernoulli diffusion model with a consensus-driven correction method, automatic segmentation and correction of blurred boundaries for thin bony structures (medial wall and floor of the orbit) in facial CT images are performed.
FSA-Net: Fractal-driven Synergistic Anatomy-aware Network for Segmenting White Line of Toldt in Laparoscopic Images
Wu, Kecheng (Hong Kong University of Science and Technology), Zhu, Lei (Hong Kong University of Science and Technology)
SegmentationConvolutional Neural NetworkTransformerImage
🎯 What it does: FSA-Net is proposed for precise segmentation of the white line (WLT) in laparoscopic images, and the first high-quality LTS (White Line of Toldt Segmentation) dataset is constructed.
FunBench: Benchmarking Fundus Reading Skills of MLLMs
Wei, Qijie (Renmin University of China), Li, Xirong (Renmin University of China)
TransformerLarge Language ModelVision Language ModelImageMultimodalityBenchmark
🎯 What it does: This paper constructs FunBench, a hierarchical fundus image benchmark based on visual question answering, to systematically evaluate the fundus reading ability of multimodal large language models (MLLMs).
Fusing Dual Encoders: Single-source Domain Generalization with Extremely Few Annotations
Wang, Ruofan (Nanjing University), Shi, Yinghuan (Southeast University)
SegmentationDomain AdaptationConvolutional Neural NetworkTransformerImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: In the field of single-source domain generalization (SDG) for medical image segmentation, the MEDU framework is proposed to enhance the model's generalization performance in scenarios with very few labeled samples.
Fusing Radiomic Features with Deep Representations for Gestational Age Estimation in Fetal Ultrasound Images
Wang, Fangyijie (University College Dublin), Silvestre, Guénolé (University College Dublin)
ClassificationExplainability and InterpretabilityConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical DataUltrasound
🎯 What it does: By integrating the deep features of fetal head ultrasound images with radiomic features, a model capable of automatically predicting gestational age is constructed.
Future Slot Prediction for Unsupervised Object Discovery in Surgical Video
Liao, Guiqiu (University of Pennsylvania), Hashimoto, Daniel A. (University of Pennsylvania)
Object DetectionSegmentationTransformerVideoBiomedical Data
🎯 What it does: This paper proposes an unsupervised object discovery framework based on dynamic slot attention and future slot prediction (DTST + slot merging), which can generate interpretable object slots in real-time and reconstruct segmentation masks in surgical videos.
FViM: Frequency Vision Mamba for Label-Free Cell Death Pathway Prediction in Lung Cancer Chemotherapy
Ye, Zhaoyi (Wuhan University), Lei, Cheng (Wuhan University)
ClassificationRecognitionExplainability and InterpretabilityDrug DiscoveryImageBiomedical Data
🎯 What it does: A label-free high-throughput cell death pathway prediction framework based on multidimensional optical time-stretch imaging flow cytometry (OTS-IFC) and frequency vision Mamba (FViM) was developed, applied to the prediction of cell death status and pathways in lung cancer chemotherapy.
GA-SAM: Geometry-Aware SAM Adaptation with Sparse Annotation-Driven Point Cloud Completion
Li, Shumeng (Nanjing University), Shi, Yinghuan (Southeast University)
SegmentationGenerationPoint CloudBiomedical DataComputed Tomography
🎯 What it does: The GA-SAM framework is proposed, which utilizes point cloud to generate global 3D shapes with only three slices of sparse annotations, and adapts the Segment Anything Model (SAM) through geometric constraints to achieve precise medical image segmentation.
Gaussian Primitive Optimized Deformable Retinal Image Registration
Tian, Xin (University of Bristol), Zhang, Hang (Cornell University)
OptimizationImageMagnetic Resonance Imaging
🎯 What it does: This paper proposes an iterative deformation registration framework based on Gaussian primitives (GPO), which achieves precise registration of retinal images by setting control nodes at significant vascular features and using Gaussian-weighted KNN to propagate displacement information.
GE2Hist: Generating Histology Images from Single-cell Gene Expression via Cross-modal Generative Network
Cai, Hongmin (South China University of Technology), Huang, Weitian (Guangdong Institute of Intelligence Science and Technology)
GenerationData SynthesisDiffusion modelAuto EncoderImageBiomedical Data
🎯 What it does: A method for generating tissue slice images using single-cell gene expression data is proposed.
GeneMorphFormer: Transformer-Driven Cross-Scale Mapping from Gene Expression to Cortical Morphology
Li, Xiao (Northwest University), Ren, Yudan (Northwest University)
TransformerBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Designed and implemented a Transformer-based GeneMorphFormer model to predict the two-dimensional coordinates of the gray/white matter boundary curve in the cerebral cortex using gene expression data.
General Methods Make Great Domain-specific Foundation Models: A Case-study on Fetal Ultrasound
Ambsdorf, Jakob (University of Copenhagen), Nielsen, Mads (University of Copenhagen)
ClassificationSegmentationTransformerContrastive LearningImageBiomedical DataUltrasound
🎯 What it does: Train a self-supervised vision transformer based on DINOv2 as a foundational model in the field of fetal ultrasound, and evaluate it on classification, segmentation, and few-shot tasks.
Generating Novel Brain Morphology by Deforming Learned Templates
Wang, Alan Q. (Stanford University), Adeli, Ehsan (Cornell Tech and Weill Cornell Medicine)
GenerationData SynthesisConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: MorphLDM is proposed, which synthesizes 3D brain MRI by learning deformable templates and generating deformation fields to capture morphological details.
Generative Unsupervised Anomaly Detection with Coarse-Fine Ensemble for Workload Reduction in 3D Non-contrast Brain CT of Emergency Room
Won, Jongjun (Asan Medical Center), Kim, Namkug (Asan Medical Center)
Anomaly DetectionDiffusion modelAuto EncoderImageBiomedical DataComputed Tomography
🎯 What it does: By using generative unsupervised anomaly detection, a coarse-fine hierarchical ensemble model (CMM and FGM) is employed to identify anomalies in brain CT scans, thereby reducing the workload in emergency radiology.
Geometric-Guided Few-Shot Dental Landmark Detection with Human-Centric Foundation Model
Wang, Anbang (Hong Kong University of Science and Technology), Li, Xiaomeng (Southern Medical University)
Object DetectionSegmentationTransformerAuto EncoderImageComputed Tomography
🎯 What it does: A framework for detecting anatomical landmarks of anterior teeth in dental CBCT images based on few-shot learning, called GeoSapiens, is proposed to address the issues of scarce labeling and high costs of manual annotation.
Geometry-Guided Local Alignment for Multi-View Visual Language Pre-Training in Mammography
Du, Yuexi (Yale University), Dvornek, Nicha C. (Yale University)
ClassificationRecognitionSegmentationTransformerVision Language ModelContrastive LearningImageTextMultimodalityMagnetic Resonance Imaging
🎯 What it does: Trained a CLIP-based breast imaging-text pre-training model GLAM, utilizing a multi-view geometry-guided local alignment mechanism to learn the correspondence between breast images and reports.
GEPAR3D: Geometry Prior-Assisted Learning for 3D Tooth Segmentation
Szczepański, Tomasz (Sano Centre for Computational Medicine), Sitek, Arkadiusz (Massachusetts General Hospital)
Object DetectionSegmentationConvolutional Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: A unified 3D tooth segmentation method GEPAR3D for instance detection and multi-class segmentation is proposed.
GL-LCM: Global-Local Latent Consistency Models for Fast High-Resolution Bone Suppression in Chest X-Ray Images
Sun, Yifei (Hangzhou Dianzi University), Ge, Ruiquan (Hangzhou Dianzi University)
SegmentationDiffusion modelAuto EncoderGenerative Adversarial NetworkImage
🎯 What it does: A bone suppression method based on the Global-Local Consistency Model (GL-LCM) is proposed, achieving rapid high-resolution bone suppression in chest X-ray images.
GLCP: Global-to-Local Connectivity Preservation for Tubular Structure Segmentation
Zhou, Feixiang (University of Liverpool), Zheng, Yalin (Ningbo Institute of Materials Technology and Engineering)
SegmentationConvolutional Neural NetworkBiomedical DataComputed Tomography
🎯 What it does: Proposes the GLCP framework, which includes Interactive Multi-head Segmentation (IMS) and Dual Attention Refinement (DAR), for simultaneously perceiving and maintaining the global connectivity and local continuity of elongated structures such as blood vessels.
GLM-SFNet: Global-Local Vision-Mamba with Semantic Fusion for Medical Image Segmentation
Chen, Jiahui (Xidian University), Liu, Kun (Beijing Institute of Technology)
SegmentationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A medical image segmentation network called GLM-SFNet is proposed, which balances global and local features to improve segmentation accuracy and boundary details.
Global and Local Contrastive Learning for Joint Representations from Cardiac MRI and ECG
Selivanov, Alexander (Technical University of Munich), Rueckert, Daniel (Technical University of Munich)
RetrievalRepresentation LearningTransformerContrastive LearningMultimodalityBiomedical DataMagnetic Resonance ImagingElectrocardiogram
🎯 What it does: By introducing global and local contrastive learning, the alignment of 12-lead electrocardiograms (ECG) with cardiac magnetic resonance imaging (CMR) is achieved, enhancing the representational capability of ECG in cardiac function assessment.
Global and Local Vision-Language Alignment for Few-Shot Learning and Few-Shot OOD Detection
Yan, Jie (Sun Yat-sen University), Wang, Ruixuan (Sun Yat-sen Univerisity)
ClassificationAnomaly DetectionTransformerLarge Language ModelVision Language ModelContrastive LearningImageBiomedical Data
🎯 What it does: This paper proposes a unified framework based on CLIP to simultaneously improve performance in few-shot classification and few-shot anomaly detection in medical images.
GoCa: Trustworthy Multi-Modal RAG with Explicit Thinking Distillation for Reliable Decision-Making in Med-LVLMs
Dai, Pengyu (Institute of Integrated Research, Institute of Science Tokyo), Suzuki, Kenji (Institute of Integrated Research, Institute of Science Tokyo)
RetrievalExplainability and InterpretabilityKnowledge DistillationTransformerLarge Language ModelContrastive LearningMultimodalityBiomedical DataRetrieval-Augmented GenerationChain-of-Thought
🎯 What it does: Proposes the GoCa multimodal retrieval-augmented generation (RAG) system, which enhances the credibility and interpretability of Med-LVLM using Chain-of-Thought (CoT) distillation and multi-agent collaboration.
GPU Accelerated Modeling of Cortical Radial and Tangential Connectivity Changes in Neurodegeneration
Zhang, Hongbo (University of Southern California), Shi, Yonggang (University of Southern California)
OptimizationDiffusion modelBiomedical DataMagnetic Resonance ImagingFibre Orientation DistributionAlzheimer's Disease
🎯 What it does: Decomposing the radial and tangential diffusion signals of the cerebral cortex, a GPU-based probabilistic optimization framework is proposed.
GradInvDiff: Stealing Medical Privacy in Federated Learning via Diffusion-Based Gradient Inversion
Wang, Zhiyuan (Beijing Institute of Technology), Liu, Kun (Beijing Institute of Technology)
Federated LearningSafty and PrivacyAdversarial AttackDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes GradInvDiff, a method for implementing gradient inversion attacks on medical images using diffusion models in a federated learning scenario.
Graph Disentanglement Learning for fMRI Analysis: Decoupling Disease, Covariates, and Individual Variability
Zhang, Shengjie (Shanghai Jiao Tong University), Zhou, Yuan (Fudan University)
ClassificationRepresentation LearningGraph Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A graph decoupling learning framework (GDL) is proposed, which decomposes the latent features of fMRI images into three parts: disease-related, covariate-related, and individual differences, and uses this framework for disease diagnosis and biomarker discovery.
Graph Laplacian Transformer with Progressive Sampling for Prostate Cancer Grading
Junayed, Masum Shah (University of Connecticut), Nabavi, Sheida (University of Connecticut)
ClassificationTransformerImageBiomedical Data
🎯 What it does: This paper proposes a prostate cancer grading system based on graph Laplacian attention Transformer and iterative refinement module, which can adaptively filter high-information patches while maintaining spatial consistency.
Graph-based Neighbor-Aware Network for Gaze-Supervised Medical Image Segmentation
Wu, Shaoxuan (Northwest University), Shen, Dinggang (Shanghai United Imaging Intelligence Co., Ltd.)
SegmentationGraph Neural NetworkContrastive LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A neighborhood-aware network based on graph neural networks (GNAN) is proposed, utilizing eye movement data to achieve weakly supervised segmentation of medical images.
Graph-PAVNet: A Graph-Based Learning Framework for Pulmonary Artery and Vein Separation Using Multimodal Feature Sampling
Li, Qingya (Northeastern University), Tan, Wenjun (Northeastern University)
SegmentationGraph Neural NetworkTransformerMultimodalityBiomedical DataComputed Tomography
🎯 What it does: For the task of separating pulmonary arteries and veins, the authors propose an end-to-end learning framework based on graph structure called Graph-PAVNet.
GRASP-PsONet: Gradient-based Removal of Spurious Patterns for PsOriasis Severity Classification
Pal, Basudha (Johnson & Johnson Innovative Medicine), Standish, Kristopher (Johnson & Johnson Innovative Medicine)
ClassificationConvolutional Neural NetworkTransformerSupervised Fine-TuningImage
🎯 What it does: A gradient tracking-based GRASP-PsONet method is proposed for the automatic identification and removal of influential samples that lead to misjudgments by the model in remote psoriasis images, thereby improving the accuracy of remote severity scoring.
GrInAdapt: Source-free Multi-Target Domain Adaptation for Retinal Vessel Segmentation
Liu, Zixuan (University of Washington), Wang, Ruikang K. (University of Washington)
SegmentationDomain AdaptationConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical Data
🎯 What it does: Proposes the GrInAdapt framework, which achieves unsupervised multi-target domain retinal vessel segmentation through multi-view image alignment, fusion, and self-supervised adaptation.
Guided Augmentation for Monocular Depth Estimation in Cell Microscopy
Viswanathan, Abhishek (Indian Institute of Technology Madras), Ramachandran, Pradeep (KLA Corporation)
Depth EstimationConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: A physics-guided enhancement strategy is proposed, utilizing EDOF images and PSF models generated by U-Net to synthesize intermediate focal plane images, thereby improving the accuracy of monocular depth estimation in cell microscopy.
Guiding Quantitative MRI Reconstruction with Phase-wise Uncertainty
Sun, Haozhong (Tsinghua University), Chen, Huijun (Tsinghua University)
Convolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a two-stage qMRI reconstruction method called PUQ, which utilizes phase-related uncertainty guidance. It first recovers multi-phase images and estimates the uncertainty of each phase through an iterative network with MC Dropout, and then performs T1/T2 mapping using pixel-level uncertainty as weights during the parameter fitting stage.
Guiding Registration with Emergent Similarity from Pre-Trained Diffusion Models
Tursynbek, Nurislam (UNC Chapel Hill), Niethammer, Marc (UCSD)
Image TranslationSegmentationConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes using the intermediate features of a pre-trained diffusion model as a similarity measure to guide the deformation registration network for medical images, achieving semantic alignment in the absence of anatomical structures.
HA-SAM: Hierarchically Adapting SAM for Nerve Segmentation in Ultrasound Images
Peng, Zihao (Huazhong University of Science and Technology), Tan, Shan (Huazhong University of Science and Technology)
SegmentationTransformerSupervised Fine-TuningImageBiomedical DataUltrasound
🎯 What it does: Designed and evaluated HA-SAM, a hierarchical adapter-based ultrasound neural segmentation method built on SAM.
HAGE: Hierarchical Alignment Gene-Enhanced Pathology Representation Learning with Spatial Transcriptomics
Dang, Thao M. (University of Texas at Arlington), Huang, Junzhou (University of Texas at Arlington)
Representation LearningConvolutional Neural NetworkContrastive LearningImageMultimodalityBiomedical Data
🎯 What it does: Proposed the HAGE framework, which utilizes gene co-expression embedding and hierarchical alignment to predict spatial transcriptomic expression from histological images.
HalF-SAM: SAM-based Haustral Fold Detection In Colonoscopy with Debris Suppression and Temporal Consistency
Golhar, Mayank (Johns Hopkins University), Durr, Nicholas J. (Johns Hopkins University)
Object DetectionSegmentationVideo
🎯 What it does: Proposed and implemented the HalF-SAM model for detecting the pyloric fold edges in colonoscopy videos.
Hallucination-Aware Multimodal Benchmark for Gastrointestinal Image Analysis with Large Vision-Language Models
Khanal, Bidur (Rochester Institute of Technology), Bhattarai, Binod (University of Aberdeen)
TransformerLarge Language ModelSupervised Fine-TuningVision Language ModelImageTextMultimodalityBenchmark
🎯 What it does: A multimodal gastrointestinal endoscopy image dataset, Gut-VLM, has been constructed, which includes diagnostic reports generated by VLM (ChatGPT-4 Omni), hallucination sentence labels annotated by medical experts, and corresponding correction texts. Based on this data, hallucination-aware fine-tuning of VLM has been conducted.
Hard Sample Mining-based Tongue Diagnosis for Fatty Liver Disease Severity Classification
Chen, Tao (Eindhoven University of Technology), Liu, Kunhong (Xiamen University)
ClassificationMixture of ExpertsImage
🎯 What it does: A framework for tongue diagnosis based on hard sample mining (HM-TDF) is proposed to accomplish multi-class classification of fatty liver severity, and the Tongue-FLD tongue image dataset is publicly released.
HARM3-Fusion: Hierarchical Attentional Representation Learning of Multi-Modal, Multi-Temporal, and Multi-Sequence Fusion for Pathological Complete Response Prediction of Head and Neck Squamous Cell Carcinoma
Wang, Jianye (Shenzhen MSU-BIT University), Zhao, Weibing (Shenzhen Msu-Bit University)
ClassificationRepresentation LearningConvolutional Neural NetworkTransformerAuto EncoderContrastive LearningMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A hierarchical attention-based multimodal fusion framework HARM-Fusion is proposed for prognostic prediction of pathological complete response (pCR) in head and neck squamous cell carcinoma (HNSCC) patients after neoadjuvant immunotherapy (NCIT).
Harnessing EHRs for Diffusion-based Anomaly Detection on Chest X-rays
Kim, Harim (Handong Global University), Hong, Charmgil (Handong Global University)
Anomaly DetectionDiffusion modelImageBiomedical DataElectronic Health Records
🎯 What it does: The study uses diffusion models combined with structured electronic health records (EHR) to achieve unsupervised anomaly detection in chest X-rays.
Harnessing Side Information for Highly Accelerated MRI
Atalık, Arda (NYU Center for Data Science), Sodickson, Daniel K. (NYU Grossman School of Medicine)
RestorationBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An end-to-end deep learning framework named TGVN is proposed, which utilizes side information (such as different contrasts from the same examination or low SNR measurements) to address the linear inverse problem of highly accelerated MRI, significantly improving reconstruction quality.
HARP: Harmonization and Adaptive Refinement of Pseudo-Labels for Cross-Domain Medical Image Segmentation
Liu, Yulong (University of Science and Technology of China), Sun, Mingzhai (University of Science and Technology of China)
SegmentationDomain AdaptationBiomedical Data
🎯 What it does: This paper proposes the HARP framework for cross-domain semi-supervised medical image segmentation, combining pseudo-label adaptive filtering and inter-domain harmonization modules.