MICCAI 2024 Papers — Page 6
International Conference on Medical Image Computing and Computer-Assisted Intervention · 856 papers
Medical Cross-Modal Prompt Hashing with Robust Noisy Correspondence Learning
Liu, Yishu (Harbin Institute of Technology), Lu, Guangming (Harbin Institute of Technology)
RetrievalTransformerLarge Language ModelPrompt EngineeringContrastive LearningMultimodalityBiomedical Data
🎯 What it does: A medical cross-modal prompt hashing (MCPH) framework is proposed to address the noise correspondence problem in medical cross-modal retrieval.
Medical Image Segmentation via Single-Source Domain Generalization with Random Amplitude Spectrum Synthesis
Qiao, Qiang (Shandong University), Guo, Qiang (Shandong University of Finance and Economics)
SegmentationDomain AdaptationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a single-source domain generalization method based on random synthesis of frequency domain amplitude spectra for medical image segmentation.
Medical Image Synthesis via Fine-Grained Image-Text Alignment and Anatomy-Pathology Prompting
Chen, Wenting (City University of Hong Kong), Li, Xiang (Nanyang Technological University)
GenerationData SynthesisLarge Language ModelPrompt EngineeringGenerative Adversarial NetworkImageTextMultimodalityComputed Tomography
🎯 What it does: A medical image synthesis framework is proposed that utilizes fine-grained image-text alignment and anatomical-pathological hints, capable of synthesizing high-quality chest X-ray images rich in anatomical and pathological information based on automatically generated detailed reports.
MediCLIP: Adapting CLIP for Few-shot Medical Image Anomaly Detection
Zhang, Ximiao (Chinese Academy of Sciences), Zhou, Xiuzhuang (University of Science and Technology of China)
SegmentationAnomaly DetectionConvolutional Neural NetworkContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A framework for unsupervised semantic segmentation based on self-supervised tasks (Gaussian Intensity Change and CutPaste) and multi-scale feature consistency is proposed, which can learn effective pixel representations in the absence of pixel-level labels.
MedMLP: An Efficient MLP-like Network for Zero-shot Retinal Image Classification
Zhou, Menghan, Zhen, Liangli (Agency for Science Technology and Research)
ClassificationComputational EfficiencyImageBiomedical Data
🎯 What it does: Designed and trained a MLP-like network MedMLP that can handle images of arbitrary resolutions, utilizing an adaptive fully connected layer AdaFC to achieve retinal image classification under zero-shot conditions.
MedSynth: Leveraging Generative Model for Healthcare Data Sharing
Kanagavelu, Renuga (Agency for Science, Technology and Research), Goh, Rick Siow Mong (Agency for Science, Technology and Research)
GenerationCompressionSafty and PrivacyTransformerGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A MedSynth framework is proposed, which utilizes an attention generator and Vision Transformer to compress large-scale medical image datasets into shareable generative models, achieving secure data sharing.
MEGFormer: enhancing speech decoding from brain activity through extended semantic representations
Boyko, Maria (Skolkovo Institute of Science and Technology), Sharaev, Maxim (Skolkovo Institute of Science and Technology)
Convolutional Neural NetworkTransformerContrastive LearningMultimodalityAudio
🎯 What it does: The study proposes the MEGFormer model, which aligns magnetoencephalography (MEG) signals with audio speech to decode perceived speech.
Memory-efficient High-resolution OCT Volume Synthesis with Cascaded Amortized Latent Diffusion Models
Huang, Kun (Nanjing University of Science and Technology), Fu, Huazhu (Agency for Science, Technology and Research (A*STAR))
SegmentationGenerationData SynthesisDiffusion modelAuto EncoderImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A Cascaded Amortized Latent Diffusion Model (CA-LDM) has been developed to achieve 512³ resolution optical coherence tomography (OCT) volume synthesis in memory-constrained environments.
MemWarp: Discontinuity-Preserving Cardiac Registration with Memorized Anatomical Filters
Zhang, Hang (Cornell University), Wang, Rongguang (University of Pennsylvania)
Image TranslationSegmentationImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A heart image registration framework called MemWarp based on memory networks is proposed, which can retain local discontinuities of anatomical boundaries without the need for segmentation masks during the inference phase.
MeshBrush: Painting the Anatomical Mesh with Neural Stylization for Endoscopy
Han, John J. (Vanderbilt University), Wu, Jie Ying (Florida International University)
Image TranslationGenerationGenerative Adversarial NetworkImageVideoMeshComputed Tomography
🎯 What it does: This paper proposes MeshBrush, a neural network-based 3D mesh shading method that learns from existing image-to-image (I2I) style transfer models to generate temporally coherent and high-fidelity videos from endoscopic renderings of patient CT scans.
MetaAD: Metabolism-Aware Anomaly Detection for Parkinson’s Disease in 3D 18F-FDG PET
Huang, Haolin (ShanghaiTech University), Wang, Qian (United Imaging Intelligence)
Anomaly DetectionConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical DataPositron Emission Tomography
🎯 What it does: This study investigates a metabolism-aware unsupervised anomaly detection framework, MetaAD, designed to highlight abnormal metabolic regions in 18F-FDG PET images of Parkinson's disease.
MetaStain: Stain-generalizable Meta-learning for Cell Segmentation and Classification with Limited Exemplars
Konwer, Aishik (Stony Brook University), Prasanna, Prateek (Stony Brook University)
ClassificationSegmentationMeta LearningConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: This paper studies a meta-learning-based staining generalization framework called MetaStain, which is used for cell segmentation and classification of histopathological images under limited samples and can quickly adapt to unseen staining conditions.
MetaUNETR: Rethinking Token Mixer Encoding for Efficient Multi-Organ Segmentation
Lyu, Pengju (Hanglok-Tech Co., Ltd.), Zhu, Jianjun (Hanglok-Tech Co., Ltd.)
SegmentationTransformerBiomedical DataComputed Tomography
🎯 What it does: Proposed the MetaUNETR model, exploring the effects of various token mixers in 3D multi-organ segmentation, and achieving efficient spatial mixing through the TriCruci layer;
MGDR: Multi-Modal Graph Disentangled Representation for Brain Disease Prediction
Jiang, Bo (Anhui University), Tang, Jin (Anhui University)
ClassificationRepresentation LearningGraph Neural NetworkContrastive LearningMultimodalityGraphBiomedical DataAlzheimer's Disease
🎯 What it does: This paper proposes a Multi-Modal Graph Disentangled Representation (MGDR) method, which uses multi-graph convolutional networks to extract features from various modalities and integrates common and private information through SVD, contrastive learning, and multi-modal perceptual attention to achieve brain disease prediction.
MH-pFLGB: Model Heterogeneous personalized Federated Learning via Global Bypass for Medical Image Analysis
Xie, Luyuan (Peking University), Wu, Zhonghai (Peking University)
ClassificationSegmentationFederated LearningImageBiomedical Data
🎯 What it does: This paper proposes a model heterogeneous personalized federated learning framework MH-pFLGB, which utilizes a global bypass model to share knowledge among medical institutions without public data, enhancing the performance of local models.
MiHATP:A Multi-Hybrid Attention Super-Resolution Network for Pathological Image Based on Transformation Pool Contrastive Learning
Xu, Zhufeng (Institute of Computing Technology Chinese Academy of Sciences), Zhao, Yi (Institute of Computing Technology Chinese Academy of Sciences)
SegmentationSuper ResolutionConvolutional Neural NetworkContrastive LearningImage
🎯 What it does: This paper proposes the MiHATP network for super-resolution of pathological images, achieving more accurate cell contour recovery through a dual-branch structure and multi-mixed attention combined with transformation pool contrastive learning.
Mining Gold from the Sand: Weakly Supervised Histological Tissue Segmentation with Activation Relocalization and Mutual Learning
Feng, Siyang (Guilin University of Electronic Technology), Pan, Xipeng (Guilin University of Electronic Technology)
SegmentationConvolutional Neural NetworkTransformerImage
🎯 What it does: This paper proposes a weakly supervised tissue segmentation framework called ARML, which enhances the quality of CAM pseudo-masks through Activation Relocalization and reduces pseudo-mask noise via mutual learning among three networks and two denoising strategies (sample-weighted voting and sample relationship mining), ultimately achieving more accurate tissue segmentation.
Misaligned 3D Texture Optimization in MIS Utilizing Generative Framework
Zheng, Jieyu (Hefei University of Technology), Ma, Xiang (Hefei University of Technology)
RestorationGenerationOptimizationTransformerGenerative Adversarial NetworkSimultaneous Localization and MappingOptical FlowImagePoint Cloud
🎯 What it does: Using camera projection to transfer 3D point cloud textures to 2D texture space, and employing a pre-trained KL-Reg VQ-GAN combined with registration and fusion modules to merge two frames of RGB point clouds, generating high-resolution continuous textures.
Misjudging the Machine: Gaze May Forecast Human-Machine Team Performance in Surgery
Cho, Sue Min (Johns Hopkins University), Unberath, Mathias (University of Arkansas)
ImageMagnetic Resonance Imaging
🎯 What it does: This study explores the performance prediction of human-machine collaboration in 2D/3D image registration evaluation through eye tracking, verifying whether eye movement count and duration can reflect evaluation errors.
Missing as Masking: Arbitrary Cross-modal Feature Reconstruction for Incomplete Multimodal Brain Tumor Segmentation
Zeng, Zhilin (Shanghai Jiao Tong University), Shen, Wei (Zhejiang University)
SegmentationTransformerMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an M FeCon model based on cross-modal feature reconstruction, which can learn feature representations that approximate complete modalities in multi-modal brain tumor segmentation, thereby improving segmentation performance under conditions of modality absence.
Mitigating attribute amplification in counterfactual image generation
Xia, Tian (Imperial College London), Glocker, Ben (Imperial College London)
GenerationData SynthesisFlow-based ModelAuto EncoderImageBiomedical DataElectronic Health Records
🎯 What it does: This study investigates the use of Deep Structural Causal Models (DSCM) to generate counterfactual images in medical imaging, identifying an attribute amplification issue and proposing a solution called Soft Counterfactual Fine-Tuning (Soft-CFT).
Mixed Integer Linear Programming for Discrete Sampling Scheme Design in Diffusion MRI
Zhang, Si-Miao (Beihang University), Cheng, Jian (Beihang University)
OptimizationDiffusion modelBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study investigates the discrete optimization problem of polarity and order in diffusion magnetic resonance imaging (dMRI) sampling schemes, proposing a global optimization method based on mixed-integer linear programming (MILP).
MM-Retinal: Knowledge-Enhanced Foundational Pretraining with Fundus Image-Text Expertise
Wu, Ruiqi (Southeast University), Fu, Huazhu (Agency for Science, Technology and Research (A*STAR))
ClassificationRecognitionSegmentationConvolutional Neural NetworkContrastive LearningImageTextMultimodality
🎯 What it does: A multimodal MM-Retinal dataset was constructed, and a knowledge-enhanced foundational model called KeepFIT was proposed for cross-task general pre-training of retinal fundus images.
MMBCD: Multimodal Breast Cancer Detection from Mammograms with Clinical History
Jain, Kshitiz (Indian Institute of Technology Delhi), Arora, Chetan (PGIMER Chandigarh)
ClassificationObject DetectionTransformerLarge Language ModelSupervised Fine-TuningImageTextMultimodality
🎯 What it does: This paper proposes a multimodal breast cancer detection model that combines mammograms with clinical history, utilizing ROI multi-instance learning and cross-modal attention fusion.
MMFusion: Multi-modality Diffusion Model for Lymph Node Metastasis Diagnosis in Esophageal Cancer
Wu, Chengyu (Shandong University), Wang, Shuai (Lishui Institute of Hangzhou Dianzi University)
ClassificationGraph Neural NetworkDiffusion modelMultimodalityBiomedical DataComputed Tomography
🎯 What it does: Combining CT images, clinical, blood, and radiomics data, the MMFusion model is proposed to fuse multimodal features using heterogeneous graphs, and to eliminate feature redundancy through a conditional feature-guided diffusion process, achieving the diagnosis of lymph node metastasis in esophageal squamous cell carcinoma.
MMQL: Multi-Question Learning for Medical Visual Question Answering
Chen, Qishen (Shanghai University), Xu, Huahu (Shanghai University)
TransformerVision Language ModelImageMultimodalityBiomedical Data
🎯 What it does: A multi-question learning (MMQL) framework is proposed for medical visual question answering tasks, which jointly processes multiple questions under the same medical image and utilizes answered questions as prompt information to enhance diagnostic accuracy.
MMSummary: Multimodal Summary Generation for Fetal Ultrasound Video
Guo, Xiaoqing (University of Oxford), Noble, J. Alison (University of Oxford)
SegmentationGenerationTransformerLarge Language ModelPrompt EngineeringVideoTextMultimodalityBiomedical DataUltrasound
🎯 What it does: The MMSummary system automatically generates multimodal summaries of fetal ultrasound videos, including keyframe detection, keyframe description generation, and anatomical segmentation measurements.
MoCo-Diff: Adaptive Conditional Prior on Diffusion Network for MRI Motion Correction
Li, Feng (ShanghaiTech University), Wang, Qian (United Imaging Intelligence)
RestorationSegmentationTransformerDiffusion modelBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A two-stage deep learning framework called MoCo-Diff is proposed, which first uses a dual-branch Transformer to remove MRI motion artifacts. The resulting motion-corrected images are then used as adaptive priors, employing a control network based on Stable Diffusion for high-quality image restoration, incorporating uncertainty prediction to achieve dynamic weighting, thereby enhancing the reliability and detail preservation of the restoration.
Modeling and Understanding Uncertainty in Medical Image Classification
Chen, Aobo (Iowa State University), Huai, Mengdi (Iowa State University)
ClassificationImageBiomedical DataComputed Tomography
🎯 What it does: An approximate full-scale conformal prediction method based on training trajectories (TAFCP) and an uncertainty explanation method (UnEX) for identifying the most influential training samples using gradient updates are proposed for medical image classification.
ModelMix: A New Model-Mixup Strategy to Minimize Vicinal Risk across Tasks for Few-scribble based Cardiac Segmentation
Zhang, Ke (Johns Hopkins University), Patel, Vishal M. (Johns Hopkins University)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes the ModelMix method for cardiac image segmentation using a small amount of scribble annotations. It constructs linear mixed virtual models of different task models and enhances segmentation performance through vicinal risk regularization.
MoME: Mixture of Multimodal Experts for Cancer Survival Prediction
Xiong, Conghao (Chinese University of Hong Kong), King, Irwin (Chinese University of Hong Kong)
ClassificationOptimizationExplainability and InterpretabilityConvolutional Neural NetworkTransformerMixture of ExpertsMultimodalityBiomedical Data
🎯 What it does: For the task of cancer survival prediction, this paper proposes a Biased Progressive Encoding (BPE) framework based on a mixture of multimodal experts, and implements a Mixture of Multimodal Experts (MoME) layer that can alternately fuse pathological Whole Slide Images (WSIs) and genomic features multiple times during the encoding process, dynamically selecting the most suitable expert to enhance prediction performance.
MoRA: LoRA Guided Multi-Modal Disease Diagnosis with Missing Modality
Shi, Zhiyi (Carnegie Mellon University), Pfister, Hanspeter (Harvard University)
ClassificationRecognitionOptimizationTransformerVision Language ModelMultimodalityBiomedical Data
🎯 What it does: This paper addresses the application of multimodal pre-training models in disease diagnosis and proposes the MoRA (Modal-aware Low-rank Adaptation) method, which solves the missing modality problem and significantly reduces training costs.
MoreStyle: Relax Low-frequency Constraint of Fourier-based Image Reconstruction in Generalizable Medical Image Segmentation
Zhao, Haoyu (Wuhan University), Xu, Yongchao (Wuhan University)
SegmentationDomain AdaptationGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A pluggable MoreStyle module is proposed, which enhances the domain generalization performance of single-source medical image segmentation by relaxing low-frequency constraints in the Fourier space, utilizing adversarial style augmentation and uncertainty-weighted loss.
MOST: Multi-Formation Soft Masking for Semi-Supervised Medical Image Segmentation
Liu, Xinyu (Chinese University of Hong Kong), Yuan, Yixuan (Southern University of Science and Technology)
SegmentationImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A multi-scale soft mask semi-supervised medical image segmentation framework called MOST is proposed.
MPMNet: Modal Prior Mutual-support Network for Age-related Macular Degeneration Classification
Li, Yuanyuan (Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences), Zhang, Jiong (Chinese Academy of Sciences)
ClassificationKnowledge DistillationConvolutional Neural NetworkTransformerImageMultimodalityComputed Tomography
🎯 What it does: A multimodal network called MPMNet is proposed, which combines OCT sequences and OCTA images to distinguish between normal, dry AMD, type 1 CNV, and type 2 CNV.
mQSM: Multitask Learning-based Quantitative Susceptibility Mapping for Iron Analysis in Brain
He, Junjie (Shanghai Jiao Tong University), Wang, Rongpin (University of Science and Technology of China)
Convolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A quantitative susceptibility mapping (QSM) reconstruction method based on frequency domain thresholding and deep learning is proposed.
MRScore: Evaluating Medical Report with LLM-based Reward System
Liu, Yunyi (University of Sydney), Zhou, Luping (University of Sydney)
TransformerLarge Language ModelReinforcement LearningTextBiomedical DataElectronic Health Records
🎯 What it does: MRScore has been developed, a medical report evaluation metric based on an LLM reward model.
MuGI: Multi-Granularity Interactions of Heterogeneous Biomedical Data for Survival Prediction
Long, Lifan (Sichuan University), Wang, Yan (East China Normal University)
ClassificationData SynthesisTransformerGenerative Adversarial NetworkMultimodalityBiomedical Data
🎯 What it does: The study proposes a multi-granularity multi-modal fusion framework MuGI to integrate pathological WSIs and genomic data to predict the survival risk of cancer patients.
Multi-category Graph Reasoning for Multi-modal Brain Tumor Segmentation
Li, Dongzhe (Guangdong University of Technology), He, Xiaochen (Guangdong University of Technology)
SegmentationConvolutional Neural NetworkGraph Neural NetworkTransformerMultimodalityBiomedical DataMagnetic Resonance ImagingBenchmark
🎯 What it does: A Multi-category Region-guided Graph Reasoning Network is proposed for multi-modal MRI brain tumor segmentation.
Multi-Dataset Multi-Task Learning for COVID-19 Prognosis
Ruffini, Filippo (Università Campus Bio-Medico di Roma), Guarrasi, Valerio (Università Campus Bio-Medico di Roma)
ClassificationConvolutional Neural NetworkImage
🎯 What it does: This study proposes a multi-dataset multi-task learning framework that combines chest X-ray images from AIforCOVID and BRIXIA, used respectively to predict the prognostic severity (mild/severe) of COVID-19 patients and the Brixia score of the lungs (four levels).
Multi-disease Detection in Retinal Images Guided by Disease Causal Estimation
Xie, Jianyang (University of Liverpool), Zheng, Yalin (Ningbo Institute of Materials Technology and Engineering)
ClassificationTransformerImage
🎯 What it does: This paper proposes a full-process multi-disease detection model based on disease causal estimation, utilizing cross-attention to obtain disease-specific features and learning a directed acyclic graph (DAG) between diseases to regularize feature learning.
Multi-Frequency and Smoke Attention-aware Learning based Diffusion Model for Removing Surgical Smoke
Li, Hao (Shanghai Jiaotong University), Huang, Pu (Shandong Normal University)
RestorationSegmentationDiffusion modelImage
🎯 What it does: A multi-frequency and smoke attention learning diffusion model is proposed to eliminate smoke during laparoscopic surgery and restore true colors.
Multi-modal Data Binding for Survival Analysis Modeling with Incomplete Data and Annotations
Qu, Linhao (Shanghai Artificial Intelligence Laboratory), Wang, Xiaosong (Shanghai Artificial Intelligence Laboratory)
ClassificationRepresentation LearningGraph Neural NetworkContrastive LearningMultimodalityBiomedical Data
🎯 What it does: A multi-modal survival analysis framework is proposed to jointly address modality missingness and right-censoring label missingness, enhancing prediction accuracy through foundational model encoding, patient-level alignment and contrastive learning, and progressive pseudo-label generation.
Multi-Modal Data Fusion with Missing Data Handling for Mild Cognitive Impairment Progression Prediction
Liu, Shuting, Rueckert, Daniel (Technical University of Munich)
ClassificationGraph Neural NetworkAuto EncoderMultimodalityBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: This paper proposes a model that integrates multimodal data and can handle missing data, aimed at predicting the progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD).
Multi-Modal Graph Neural Network with Transformer-Guided Adaptive Diffusion for Preclinical Alzheimer Classification
Sim, Jaeyoon (Pohang University of Science and Technology), Kim, Won Hwa (Pohang University of Science and Technology)
ClassificationExplainability and InterpretabilityGraph Neural NetworkTransformerDiffusion modelMultimodalityBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyDiffusion Tensor ImagingAlzheimer's Disease
🎯 What it does: A multi-modal graph neural network (GTAD) is proposed, which achieves early Alzheimer's disease classification on brain networks through Transformer-guided adaptive diffusion.
Multi-modality 3D CNN Transformer for Assisting Clinical Decision in Intracerebral Hemorrhage
Xiong, Zicheng (Henan University), Yang, Fuxing (Fujian Medical University)
Convolutional Neural NetworkTransformerImageMultimodalityTabularBiomedical DataComputed TomographyElectronic Health Records
🎯 What it does: A multimodal 3D CNN-Transformer model that combines CT images and clinical tabular data is proposed for surgical or conservative treatment decision-making in early-stage patients with cerebral hemorrhage.
Multi-order Simplex-based Graph Neural Network for Brain Network Analysis
Hwang, Yechan (Pohang University of Science and Technology), Kim, Won Hwa (Pohang University of Science and Technology)
ClassificationExplainability and InterpretabilityGraph Neural NetworkSupervised Fine-TuningMultimodalityBiomedical DataPositron Emission TomographyAlzheimer's Disease
🎯 What it does: A multi-order simplex graph neural network is proposed, which can explicitly learn the embeddings of nodes and edges in brain networks and capture their interrelationships through a dual aggregation framework.
Multi-scale Region-aware Implicit Neural Network for Medical Images Matting
Xu, Yanyu (Shandong University), Xu, Xinxing (Agency for Science, Technology and Research)
SegmentationGenerationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a multi-scale region-aware implicit neural network for generating soft masks (alpha matte) in medical images, helping to address the resolution issues of traditional binary segmentation in fuzzy lesion areas.
Multi-sequence learning for multiple sclerosis lesion segmentation in spinal cord MRI
Walsh, Ricky (Univ Rennes), Galassi, Francesca (Univ Rennes)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study investigates the use of multi-sequence learning for the segmentation of spinal cord lesions in multiple sclerosis and proposes a complete preprocessing and multi-modal U-Net model scheme.
Multi-stage Multi-granularity Focus-tuned Learning Paradigm for Medical HSI Segmentation
Dong, Haichuan (East China Normal University), Wang, Yan (East China Normal University)
SegmentationConvolutional Neural NetworkBiomedical Data
🎯 What it does: A multi-stage, multi-granularity Focus-tuned Learning framework is proposed for medical hyperspectral image segmentation, which can simultaneously focus on subtle spectral differences at both the disease and image levels while balancing spatial-spectral features.
Multilevel Causality Learning for Multi-label Gastric Atrophy Diagnosis
Cui, Xiaoxiao (Shandong University), Li, Shuo (Harbin Institute of Technology)
ClassificationTransformerImageBiomedical Data
🎯 What it does: A multi-layer causal learning framework is proposed for multi-label gastric atrophy diagnosis, capable of simultaneously detecting atrophy and gastric location.
Multimodal Cross-Task Interaction for Survival Analysis in Whole Slide Pathological Images
Jiang, Songhan (Harbin Institute of Technology (Shenzhen)), Zhang, Yongbing (Harbin Institute of Technology Shenzhen)
ClassificationRepresentation LearningTransformerSupervised Fine-TuningImageMultimodalityBiomedical Data
🎯 What it does: A multi-modal cross-task interaction framework MCTI based on subclass classification assistance is proposed to integrate pathological WSI and gene expression for tumor survival prediction.
Multimodal Learning for Embryo Viability Prediction in Clinical IVF
Kim, Junsik (Harvard University), Pfister, Hanspeter (Harvard University)
ClassificationSegmentationAnomaly DetectionTransformerVideoMultimodalityTabularBiomedical DataElectronic Health Records
🎯 What it does: This paper develops a multimodal prediction model that integrates delayed embryo time-lapse videos with patients' electronic health records (EHR) to assess the viability of IVF embryos and predict pregnancy success rates.
Multimodal Variational Autoencoder for Low-cost Cardiac Hemodynamics Instability Detection
Suvon, Mohammod N. I. (University of Sheffield), Lu, Haiping (University of Sheffield)
Anomaly DetectionConvolutional Neural NetworkAuto EncoderMultimodalityBiomedical DataElectronic Health RecordsElectrocardiogram
🎯 What it does: This paper proposes a multimodal variational autoencoder (CardioVAE X,G) that utilizes low-cost chest X-ray and electrocardiogram data for predicting cardiovascular hemodynamic instability (CHDI), particularly the estimation of pulmonary artery wedge pressure (PAWP);
Multivariate Cooperative Game for Image-Report Pairs: Hierarchical Semantic Alignment for Medical Report Generation
Zhu, Zhihong (Peking University), Zheng, Yefeng (Westlake University)
GenerationTransformerContrastive LearningImageTextMultimodalityElectronic Health Records
🎯 What it does: This paper proposes a multi-level semantic alignment framework (HSA) that maps image-report pairs to different levels of players, achieving cross-modal semantic alignment at the case level, region level, and disease level, thereby improving the quality of medical report generation.
MultiVarNet - Predicting Tumour Mutational status at the Protein Level
Morel, Louis-Oscar (University of Oxford), Rittscher, Jens (University of Oxford)
ClassificationSegmentationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A deep learning framework named MultiVarNet was developed and validated, utilizing H&E stained images to predict tumor mutation variants at the protein level, surpassing traditional gene-level predictions.
MuST: Multi-Scale Transformers for Surgical Phase Recognition
Pérez, Alejandra (Universidad de los Andes), Arbeláez, Pablo (Universidad de los Andes)
RecognitionTransformerVideo
🎯 What it does: A multi-scale Transformer-based surgical phase recognition framework called MuST is proposed, which utilizes a multi-level temporal pyramid and cross-scale attention to achieve joint modeling of short-term, mid-term, and long-term information in surgical videos.
Myocardial Scar Enhancement in LGE Cardiac MRI using Localized Diffusion
Hasny, Marta (Beth Israel Deaconess Medical Center and Harvard Medical School), Nezafat, Reza (Beth Israel Deaconess Medical Center and Harvard Medical School)
RestorationSegmentationConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study proposes a local conditional diffusion model to enhance the contrast between scars and normal myocardium in LGE cardiac magnetic resonance imaging.
Neural Cellular Automata for Lightweight, Robust and Explainable Classification of White Blood Cell Images
Deutges, Michael (Helmholtz Zentrum München - German Research Center for Environmental Health), Marr, Carsten (Helmholtz Zentrum München - German Research Center for Environmental Health)
ClassificationExplainability and InterpretabilityImage
🎯 What it does: A lightweight, robust, and interpretable single-cell leukocyte classification method based on Neural Cellular Automata (NCA) is proposed and validated on three datasets from different sources.
NeuroConText: Contrastive Text-to-Brain Mapping for Neuroscientific Literature
Meudec, Raphaël (Université Paris-Saclay), Thirion, Bertrand (Université Paris-Saclay)
RetrievalTransformerLarge Language ModelContrastive LearningTextBiomedical Data
🎯 What it does: NeuroConText is proposed, a brain-text mapping framework based on contrastive learning that can associate the text of neuroscience articles with brain activation coordinates and generate brain activation maps from text.
NeuroLink: Bridging Weak Signals in Neuronal Imaging with Morphology Learning
Yan, Haiyang (Chinese Academy of Sciences), Han, Hua (Chinese Academy of Sciences)
RestorationSegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: NeuroLink is proposed, a 3D neuron segmentation and reconstruction method that combines dynamic snake convolution, weighted local connectivity loss, and multi-task learning, addressing the weak signal problem in low-contrast images.
nnU-Net Revisited: A Call for Rigorous Validation in 3D Medical Image Segmentation
Isensee, Fabian (DKFZ), Jäger, Paul F. (German Cancer Research Center)
SegmentationConvolutional Neural NetworkTransformerBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Through systematic evaluation, validation trap identification, and comparative analysis, the current state of 3D medical image segmentation is re-examined, and more rigorous validation standards are proposed.
No-New-Denoiser: A Critical Analysis of Diffusion Models for Medical Image Denoising
Pfaff, Laura (FAU Erlangen-Nrnberg), Maier, Andreas (Friedrich-Alexander-Universität Erlangen)
RestorationDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The paper systematically evaluates the performance of diffusion models in medical image denoising, particularly the challenges in maintaining the integrity of image content.
NODER: Image Sequence Regression Based on Neural Ordinary Differential Equations
Bai, Hao (Shanghai Jiao Tong University), Hong, Yi (Shanghai Jiao Tong University)
RestorationOptimizationAuto EncoderImageBiomedical DataMagnetic Resonance ImagingAlzheimer's DiseaseOrdinary Differential Equation
🎯 What it does: This paper studies a 3D medical image sequence regression framework based on Neural Ordinary Differential Equations (NODER) to predict missing or future time point images.
Noise Level Adaptive Diffusion Model for Robust Reconstruction of Accelerated MRI
Huang, Shoujin (Shenzhen Technology University), Lyu, Mengye (Longgang Central Hospital of Shenzhen)
RestorationDiffusion modelImageMagnetic Resonance Imaging
🎯 What it does: A noise level adaptive diffusion model (Nila-DC) is proposed to accelerate MRI image reconstruction, capable of maintaining high-quality reconstruction in the presence of measurement noise.
Noise Removed Inconsistency Activation Map for Unsupervised Registration of Brain Tumor MRI between Pre-operative and Follow-up Phases
Wu, Chongwei (Huazhong University of Science and Technology), Wang, Zhiwei (Huazhong University of Science and Technology)
ImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an unsupervised image registration method for preoperative and follow-up MRI of brain tumors, which improves registration accuracy by generating and utilizing a noise-reduced structural inconsistency activation map (NR-IAM) to locate and correct structurally inconsistent areas.
Non-Adversarial Learning: Vector-Quantized Common Latent Space for Multi-Sequence MRI
Han, Luyi (Radboud University Medical Centre), Mann, Ritse (Nanjing University of Information Science and Technology)
SegmentationGenerationData SynthesisAuto EncoderContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A non-adversarial generative framework based on VQ-VAE and dynamic Seq2Seq (VQ-Seq2Seq) is proposed, which constructs a common latent representation for multi-sequence MRI using a discretized latent space, thereby enabling the synthesis of missing sequences.
Nonrigid Reconstruction of Freehand Ultrasound without a Tracker
Li, Qi (University College London), Hu, Yipeng (King's College London)
Image TranslationOptimizationMeta LearningConvolutional Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: A framework for freehand ultrasound (US) three-dimensional reconstruction without external trackers is proposed, jointly estimating the rigid motion of the probe and the non-rigid deformation of soft tissues.
Novelty Detection Based Discriminative Multiple Instance Feature Mining to Classify NSCLC PD-L1 Status on HE-Stained Histopathological Images
Xu, Rui (Dalian University of Technology), Hu, Hongjie (Zhejiang University)
ClassificationAnomaly DetectionConvolutional Neural NetworkContrastive LearningImage
🎯 What it does: A novel detection-based multi-instance feature mining method is proposed to predict the PD-L1 status of NSCLC using HE-stained tissue sections, replacing expensive IHC testing.
OCL: Ordinal Contrastive Learning for Imputating Features with Progressive Labels
Baek, Seunghun (Pohang University of Science and Technology), Kim, Won Hwa (Pohang University of Science and Technology)
ClassificationRestorationAuto EncoderContrastive LearningMultimodalityBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease
🎯 What it does: A global feature completion framework based on an encoder-decoder is proposed, utilizing existing imaging modalities to infer the missing modality.
Ocular Stethoscope: Auditory Support for Retinal Membrane Peeling
Matinfar, Sasan (Technische Universität München), Navab, Nassir (Technische Universität München)
RecognitionSegmentationImageBiomedical DataComputed TomographyAudio
🎯 What it does: This paper proposes an unsupervised acoustic mapping method that sonifies the epiretinal membrane (ERM) and the slight elevation of the retinal structure through real-time OCT A-scans, assisting surgeons in identifying suitable starting points during peeling surgery.
On Instabilities of Unsupervised Denoising Diffusion Models in Magnetic Resonance Imaging Reconstruction
Han, Tianyu (University Hospital Aachen), Truhn, Daniel (University Hospital Aachen)
RestorationAdversarial AttackDiffusion modelImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: This paper studies the vulnerability of unsupervised denoising diffusion models to worst-case adversarial perturbations in magnetic resonance imaging reconstruction.
On predicting 3D bone locations inside the human body
Dakri, Abdelmouttaleb (Univ. Grenoble Alpes), Pujades, Sergi (Max Planck Institute for Intelligent Systems)
SegmentationPose EstimationPoint CloudMeshBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a method for predicting the three-dimensional positions of internal human bones using external body surface observations. First, the bone binary segmentation in the HIT dataset is refined to obtain point clouds of five long bones (humerus, radius-ulna, pelvis, femur, tibia-fibula). Then, by extending the SKEL model to SKEL-J and adding joint offset parameters ΔJ, precise registration of both skin and bones is achieved, ultimately learning the mapping from the skin surface to the bone positions.
On-the-Fly Guidance Training for Medical Image Registration
Xin, Yuelin, Xie, Xiaohui (University of California, Irvine)
OptimizationConvolutional Neural NetworkTransformerImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A novel On-the-Fly Guidance (OFG) training framework is proposed, which utilizes a differentiable optimizer to generate pseudo-labels in real-time during the training process, elevating the learning-based image registration model from unsupervised to near-supervised status.
One registration is worth two segmentations
Huang, Shiqi (University College London), Hu, Yipeng (King's College London)
SegmentationOptimizationContrastive LearningImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: This paper proposes to reformulate image registration as a two-stage multi-class segmentation problem that seeks corresponding ROIs in two images, and accomplishes registration in an unsupervised manner.
Online 3D reconstruction and dense tracking in endoscopic videos
Hayoz, Michel (ARTORG Center, University of Bern), Sznitman, Raphael (ARTORG Center, University of Bern)
Object TrackingSegmentationDepth EstimationGaussian SplattingOptical FlowImageVideo
🎯 What it does: Developed an online dense 3D reconstruction and tracking framework suitable for binocular endoscopic videos, capable of real-time construction and tracking of soft tissue surfaces.
Online learning in motion modeling for intra-interventional image sequences
Gunnarsson, Niklas (Uppsala University), Schön, Thomas B. (Uppsala University)
SegmentationGenerationOptimizationGaussian SplattingOptical FlowImageBiomedical DataMagnetic Resonance ImagingUltrasound
🎯 What it does: This paper proposes a differential motion model based on a low-dimensional linear Gaussian state space model, capable of estimating and predicting motion in continuous medical image sequences.
Open-Set Semi-Supervised Medical Image Classification with Learnable Prototypes and Outlier Filter
He, Along (Nankai University), Fu, Huazhu (Agency for Science, Technology and Research)
ClassificationAnomaly DetectionTransformerSupervised Fine-TuningImageBiomedical Data
🎯 What it does: Proposes the OpenSSC framework for open-set semi-supervised classification of medical images, which includes learnable prototypes, multiple binary discriminators, and a joint anomaly filter.
Optimizing Efficiency and Effectiveness in Sequential Prompt Strategy for SAM using Reinforcement Learning
Huang, Yifei (East China Normal University), Cai, Haibin (East China Normal University)
SegmentationOptimizationReinforcement LearningPrompt EngineeringImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: Proposes the AIES mechanism, which utilizes reinforcement learning to optimize the interactive prompting format and early stopping decisions of the Segment Anything Model (SAM), thereby improving the efficiency and accuracy of medical image segmentation.
ORacle: Large Vision-Language Models for Knowledge-Guided Holistic OR Domain Modeling
Özsoy, Ege (Technische Universität München), Navab, Nassir (Technische Universität München)
SegmentationGenerationData SynthesisTransformerLarge Language ModelVision Language ModelDiffusion modelImageMultimodality
🎯 What it does: A system called ORacle based on a large visual language model (LVLM) is proposed, capable of generating semantic scene graphs of operating rooms directly from multi-view RGB images, achieving global operating room modeling.
ORCGT: Ollivier-Ricci Curvature-based Graph Model for Lung STAS Prediction
Cen, Min (University of Science and Technology of China), Wang, Liansheng (Shanghai Changhai Hospital)
ClassificationSegmentationGraph Neural NetworkImageBiomedical Data
🎯 What it does: This paper proposes a graph model based on Ollivier-Ricci curvature, which first extracts circular regions from rough tumor edge annotations and then uses graph neural networks to model the cells within that region, thereby automatically predicting STAS in lung adenocarcinoma slices.
Ordinal Learning: Longitudinal Attention Alignment Model for Predicting Time to Future Breast Cancer Events from Mammograms
Wang, Xin (Chengdu University of Information Technology), Mann, Ritse (Nanjing University of Information Science and Technology)
ClassificationExplainability and InterpretabilityConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A longitudinal mammogram prediction model named OA-BreaCR is proposed, utilizing ordinal learning and attention alignment to simultaneously predict the risk of breast cancer and its occurrence time.
OSAL-ND: Open-set Active Learning for Nucleus Detection
Tang, Jiao (Nanjing University of Aeronautics and Astronautics), Shao, Wei (Nanjing University of Aeronautics and Astronautics)
Object DetectionConvolutional Neural NetworkImage
🎯 What it does: This paper proposes a two-stage open-set active learning framework OSAL-ND, aimed at significantly reducing the annotation cost of pathological image nucleus detection.
Overcoming Atlas Heterogeneity in Federated Learning for Cross-site Connectome-based Predictive Modeling
Liang, Qinghao (Yale University), Scheinost, Dustin (Yale University)
Federated LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Achieve collaborative training of connectomes under different brain region atlases through federated learning at decentralized medical imaging centers, proposing the FLECHA framework to aggregate models after mapping connectomes to a unified space locally.
Overlay Mantle-Free for Semi-Supervised Medical Image Segmentation
Liu, Jiacheng (Yunnan University), Liu, Peng (Yunnan University)
SegmentationKnowledge DistillationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A two-stage knowledge distillation framework is proposed, combining real-label-based Overlay edge cropping and stitching data augmentation for semi-supervised medical image segmentation, particularly targeting left atrium (LA) MRI data.
Pair Shuffle Consistency for Semi-supervised Medical Image Segmentation
He, Jianjun (Sun Yat-sen University), Ma, Andy J. (Sun Yat-sen University)
SegmentationKnowledge DistillationImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A semi-supervised medical image segmentation method based on pair shuffling is proposed, which splits labeled and unlabeled images into small patches and randomly rearranges them during training to enhance local information learning.
PAMIL: Prototype Attention-based Multiple Instance Learning for Whole Slide Image Classification
Liu, Jiashuai (Xi'an Jiaotong University), Gao, Zeyu (University of Cambridge)
ClassificationExplainability and InterpretabilityImage
🎯 What it does: A Prototype Attention-based Multiple Instance Learning (PAMIL) model is proposed for multi-label classification of whole slide images (WSI), providing instance-level and prototype-level explanations through a dual-branch prototype and attention mechanism.
PANS: Probabilistic Airway Navigation System for Real-time Robust Bronchoscope Localization
Tian, Qingyao (Institute of Automation, Chinese Academy of Sciences), Liu, Hongbin (Centre for Artificial Intelligence and Robotics, Hong Kong Institute of Science Innovation, Chinese Academy of Sciences)
Object DetectionPose EstimationDepth EstimationGenerative Adversarial NetworkOptical FlowImageMultimodalityBiomedical DataComputed Tomography
🎯 What it does: A real-time bronchoscope localization system called PANS is proposed based on a probabilistic Monte Carlo framework, achieving six degrees of freedom localization through depth estimation and semantic analysis.
Parameter Efficient Fine Tuning for Multi-scanner PET to PET Reconstruction
Kim, Yumin (Yonsei University), Hwang, Seong Jae (Mediwhale)
RestorationTransformerGenerative Adversarial NetworkBiomedical DataPositron Emission TomographyAlzheimer's Disease
🎯 What it does: Under various PET scanners, a method called PETITE is proposed to achieve the reconstruction from short-term PET scans to long-term scans using Parameter-Efficient Fine-Tuning (PEFT);
PASSION for Dermatology: Bridging the Diversity Gap with Pigmented Skin Images from Sub-Saharan Africa
Gottfrois, Philippe (University Hospital Basel), Navarini, Alexander A. (University of Basel)
ClassificationDomain AdaptationConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: What was done: A high-quality image dataset of skin diseases in children from sub-Saharan Africa, called PASSION, was collected, and a baseline AI diagnostic model was provided.
PASTA: Pathology-Aware MRI to PET CroSs-modal TrAnslation with Diffusion Models
Li, Yitong (Technical University of Munich), Wachinger, Christian (Technical University of Munich)
Image TranslationGenerationData SynthesisConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease
🎯 What it does: The PASTA framework is proposed, utilizing conditional diffusion models to convert structural magnetic resonance imaging (MRI) into functional positron emission tomography (PET) images, generating synthetic PET that retains pathological information related to Alzheimer's disease.
Patch-Slide Discriminative Joint Learning for Weakly-Supervised Whole Slide Image Representation and Classification
Yu, Jiahui (Zhejiang University), Xu, Yingke (University of British Columbia)
ClassificationRepresentation LearningTransformerContrastive LearningImageBiomedical Data
🎯 What it does: The PSJA-MIL framework is proposed, which jointly optimizes patch and whole slide image (WSI) level features. It enhances the discriminative power of weakly supervised WSI classification through patch contrastive loss estimation, adaptive cross-entropy loss, and a prototype cosine similarity-based PCS-classifier.
PathM3: A Multimodal Multi-Task Multiple Instance Learning Framework for Whole Slide Image Classification and Captioning
Zhou, Qifeng (University of Texas at Arlington), Huang, Junzhou (University of Texas at Arlington)
ClassificationGenerationTransformerLarge Language ModelImageTextMultimodality
🎯 What it does: A multi-modal, multi-task, multi-instance learning framework (PathM3) is proposed to achieve joint learning and inference of whole slide images (WSI) and their diagnostic captions, balancing classification and automatic generation of diagnostic captions.
PathMamba: Weakly Supervised State Space Model for Multi-class Segmentation of Pathology Images
Fan, Jiansong (Jiangnan University), Pan, Xiang (Hangzhou Dianzi University)
SegmentationContrastive LearningImage
🎯 What it does: A weakly supervised multi-class pathological image segmentation framework called PathMamba is proposed, which can learn segmentation masks from pixel-level and block-level features using only image-level labels.
Pathological Semantics-Preserving Learning for H&E-to-IHC Virtual Staining
Chen, Fuqiang (Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences), Qin, Wenjian (Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences)
Image TranslationGenerationData SynthesisConvolutional Neural NetworkGenerative Adversarial NetworkImage
🎯 What it does: A new H&E-to-IHC virtual staining method called PSPStain is proposed, which can directly generate IHC images from H&E images while preserving molecular-level protein expression information and addressing spatial mismatch issues.
PathoTune: Adapting Visual Foundation Model to Pathological Specialists
Lu, Jiaxuan, Zhang, Shaoting (Zhongda Hospital)
Domain AdaptationTransformerPrompt EngineeringMultimodalityBiomedical Data
🎯 What it does: A multi-modal prompt tuning framework named PathoTune is proposed to quickly transfer visual or pathological base models to various pathological diagnosis tasks.
Patient-Specific Real-Time Segmentation in Trackerless Brain Ultrasound
Dorent, Reuben (Harvard Medical School, Brigham and Women's Hospital), Wells III, William M. (Harvard Medical School)
SegmentationData SynthesisConvolutional Neural NetworkAuto EncoderImageBiomedical DataMagnetic Resonance ImagingUltrasound
🎯 What it does: This study presents the first patient-specific, real-time, non-tracking brain ultrasound segmentation framework, which generates virtual ultrasound scans using preoperative MRI and synthesizes intraoperative ultrasound (iUS), followed by training a 2D UNet on the synthesized data for brain tumor target segmentation.
PEMMA: Parameter-Efficient Multi-Modal Adaptation for Medical Image Segmentation
Saadi, Nada (Mohamed bin Zayed University of Artificial Intelligence), Nandakumar, Karthik (Mohamed Bin Zayed University of Artificial Intelligence)
SegmentationTransformerPrompt EngineeringMultimodalityBiomedical DataComputed TomographyPositron Emission Tomography
🎯 What it does: A parameter-efficient multimodal adaptation framework PEMMA based on LoRA is proposed, which can lightweight upgrade the UNETR model trained only on CT to a bimodal model that simultaneously utilizes PET data.
PEPSI: Pathology-Enhanced Pulse-Sequence-Invariant Representations for Brain MRI
Liu, Peirong (Harvard Medical School and Massachusetts General Hospital), Iglesias, Juan E. (Harvard Medical School and Massachusetts General Hospital)
SegmentationGenerationData SynthesisConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: We propose PEPSI, a pulse sequence invariant feature learning model based on synthetic lesion encoding, for brain MRI image synthesis and lesion segmentation.
Perspective+ Unet: Enhancing Segmentation with Bi-Path Fusion and Efficient Non-Local Attention for Superior Receptive Fields
Hu, Jintong (Tsinghua University), Yang, Wenming (Tsinghua University)
SegmentationConvolutional Neural NetworkTransformerBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A Perspective+Unet framework is proposed, combining dual-path encoding, efficient non-local Transformer blocks, and cross-scale fusion to enhance 3D medical image segmentation.