【深度学习】CVPR 2024医学影像AI相关论文！-技术圈

转自：第一作者EDesk

2 0 24

● CVPR (CCF-A).

● 来源：https://github.com/MedAIerHHL/CVPR-MIA （持续更新，已授权）

CVPR-MIA

Image Reconstruction (图像重建)

QN-Mixer: A Quasi-Newton MLP-Mixer Model for Sparse-View CT Reconstruction.

中文：QN-Mixer：用于稀疏视图CT重建的拟牛顿MLP-Mixer模型
Paper: https://arxiv.org/abs/2402.17951v1
Project: https://towzeur.github.io/QN-Mixer/

Fully Convolutional Slice-to-Volume Reconstruction for Single-Stack MRI.

中文：单栈MRI的全卷积切片到体积重建
Paper: https://arxiv.org/abs/2312.03102
Code: http://github.com/seannz/svr

Structure-Aware Sparse-View X-ray 3D Reconstruction.

中文：结构感知稀疏视图 X 射线 3D 重建
Paper: https://arxiv.org/abs/2311.10959
Code: https://github.com/caiyuanhao1998/SAX-NeRF

Progressive Divide-and-Conquer via Subsampling Decomposition for Accelerated MRI.

中文：通过子采样分解的渐进分治以加速MRI
Paper: https://arxiv.org/abs/2403.10064
Code: https://github.com/ChongWang1024/PDAC

Image Resolution (图像超分)

Learning Large-Factor EM Image Super-Resolution with Generative Priors

中文：使用生成先验学习大因子电磁图像超分辨率
Paper: https://openaccess.thecvf.com/content/CVPR2024/papers/Shou_Learning_Large-Factor_EM_Image_Super-Resolution_with_Generative_Priors_CVPR_2024_paper.pdf
Code: https://github.com/jtshou/GPEMSR
Video: https://youtu.be/LNSLQM5-YcM

CycleINR: Cycle Implicit Neural Representation for Arbitrary-Scale Volumetric Super-Resolution of Medical Data

中文：CycleINR：任意尺度医学数据体素超分辨率的循环隐式神经表示
Paper: https://arxiv.org/abs/2404.04878v1

Image Registration (图像配准)

Modality-Agnostic Structural Image Representation Learning for Deformable Multi-Modality Medical Image Registration

中文：适用于可变形多模态医学图像配准的模态无关结构图像表示学习
Paper: https://arxiv.org/abs/2402.18933

[Oral & Best Paper Candidate!!!] Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registration

中文：基于相关性的粗到细MLP用于可变形医学图像配准
Paper: https://arxiv.org/abs/2406.00123
Code: https://github.com/jungeun122333/UVI-Net

Image Segmentation (图像分割)

PrPSeg: Universal Proposition Learning for Panoramic Renal Pathology Segmentation

中文：PrPSeg：全景肾病病理分割的通用命题学习
Paper: https://arxiv.org/abs/2402.19286

Versatile Medical Image Segmentation Learned from Multi-Source Datasets via Model Self-Disambiguation

中文：通过模型自我消歧学习的多功能医学图像分割，来自多源数据集
Paper: https://arxiv.org/abs/2311.10696

Each Test Image Deserves A Specific Prompt: Continual Test-Time Adaptation for 2D Medical Image Segmentation

中文：每个测试图像应得到特定提示：2D医学图像分割的持续测试时适应
Paper: https://arxiv.org/abs/2311.18363
Code: https://github.com/Chen-Ziyang/VPTTA

One-Prompt to Segment All Medical Images

中文：一提示分割所有医学图像
Paper: https://arxiv.org/abs/2305.10300
Code: https://github.com/WuJunde/PromptUNet/tree/main

Modality-agnostic Domain Generalizable Medical Image Segmentation by Multi-Frequency in Multi-Scale Attention

中文：基于多尺度注意力的多频率模态无关医学图像分割
Paper: https://arxiv.org/abs/2405.06284
Code Project: https://skawngus1111.github.io/MADGNet_project/

Diversified and Personalized Multi-rater Medical Image Segmentation

中文：多样化和个性化的多评分员医学图像分割
Paper: https://arxiv.org/pdf/2212.00601
Code: https://github.com/ycwu1997/D-Persona

MAPSeg: Unified Unsupervised Domain Adaptation for Heterogeneous Medical Image Segmentation Based on 3D Masked Autoencoding and Pseudo-Labeling

中文：基于3D遮罩自动编码和伪标签的MAPSeg：异构医学图像分割的统一无监督域适应
Paper: https://arxiv.org/abs/2303.09373

Adaptive Bidirectional Displacement for Semi-Supervised Medical Image Segmentation

中文：半监督医学图像分割的自适应双向位移
Paper: https://arxiv.org/abs/2405.00378
Code: https://github.com/chy-upc/ABD

Cross-dimension Affinity Distillation for 3D EM Neuron Segmentation

中文：3D EM神经元分割的跨维度亲和力蒸馏
Paper: https://openaccess.thecvf.com/content/CVPR2024/papers/Liu_Cross-Dimension_Affinity_Distillation_for_3D_EM_Neuron_Segmentation_CVPR_2024_paper.pdf
Code: https://github.com/liuxy1103/CAD

ToNNO: Tomographic Reconstruction of a Neural Network’s Output for Weakly Supervised Segmentation of 3D Medical Images.

中文：ToNNO：神经网络输出的断层重建用于弱监督3D医学图像分割
Paper: https://arxiv.org/abs/2405.06880

Versatile Medical Image Segmentation Learned from Multi-Source Datasets via Model Self-Disambiguation

中文：通过模型自我消歧学习的多功能医学图像分割，来自多源数据集
Paper: https://arxiv.org/abs/2311.10696

Teeth-SEG: An Efficient Instance Segmentation Framework for Orthodontic Treatment based on Anthropic Prior Knowledge

中文：基于人类先验知识的正畸治疗高效实例分割框架
Paper: https://arxiv.org/abs/2404.01013

Tyche: Stochastic in Context Learning for Universal Medical Image Segmentation

中文：Tyche：上下文中的随机学习用于通用医学图像分割
Paper: https://arxiv.org/abs/2401.13650
Code: https://github.com/mariannerakic/tyche/

Constructing and Exploring Intermediate Domains in Mixed Domain Semi-supervised Medical Image Segmentation

中文：混合域半监督医学图像分割中中间域的构建与探索
Paper: https://arxiv.org/abs/2404.08951
Code: https://github.com/MQinghe/MiDSS

S2VNet: Universal Multi-Class Medical Image Segmentation via Clustering-based Slice-to-Volume Propagation

中文：S2VNet：通过聚类基础的切片到体积传播实现通用多类别医学图像分割
Paper: https://arxiv.org/abs/2403.16646
Code: https://github.com/dyh127/S2VNet

EMCAD: Efficient Multi-scale Convolutional Attention Decoding for Medical Image Segmentation.

中文：EMCAD：医学图像分割的高效多尺度卷积注意力解码
Paper: https://arxiv.org/abs/2405.06880
Code: https://github.com/SLDGroup/EMCAD

Training Like a Medical Resident: Context-Prior Learning Toward Universal Medical Image Segmentation.

中文：像住院医生一样训练：情境先验学习导向的通用医学图像分割
Paper: https://arxiv.org/abs/2306.02416
Code: https://github.com/yhygao/universal-medical-image-segmentation

ZePT: Zero-Shot Pan-Tumor Segmentation via Query-Disentangling and Self-Prompting

Paper: https://arxiv.org/abs/2312.04964

[Oral!!!] Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registration

Paper: https://github.com/dengxl0520/MemSAM/blob/main/paper.pdf
Code: https://github.com/dengxl0520/MemSAM/tree/main

PH-Net: Semi-Supervised Breast Lesion Segmentation via Patch-wise Hardness

Paper: https://openaccess.thecvf.com/content/CVPR2024/papers/Jiang_PH-Net_Semi-Supervised_Breast_Lesion_Segmentation_via_Patch-wise_Hardness_CVPR_2024_paper.pdf
Code: https://github.com/jjjsyyy/PH-Net
Video: https://cvpr.thecvf.com/virtual/2024/poster/30539

Image Generation (图像生成)

Learned representation-guided diffusion models for large-image generation

中文：用于大型图像生成的学习表示指导的扩散模型
Paper: https://arxiv.org/abs/2312.07330

MedM2G: Unifying Medical Multi-Modal Generation via Cross-Guided Diffusion with Visual Invariant

中文：MedM2G：通过视觉不变的交叉引导扩散统一医疗多模式生成
Paper: https://arxiv.org/html/2403.04290v1

Towards Generalizable Tumor Synthesis

中文：迈向泛化肿瘤合成
Paper: https://arxiv.org/abs/2402.19470v1
Code: https://github.com/MrGiovanni/DiffTumor

Data-Efficient Unsupervised Interpolation Without Any Intermediate Frame for 4D Medical Images

中文：无需任何中间帧的4D医学图像的数据高效无监督插值
Paper: https://arxiv.org/abs/2404.01464
Code: https://github.com/jungeun122333/UVI-Net

Image Classification (图像分类)

Systematic comparison of semi-supervised and self-supervised learning for medical image classification

中文：医学图像分类中半监督和自监督学习的系统比较
Paper: https://arxiv.org/abs/2307.08919v2
Code: https://github.com/tufts-ml/SSL-vs-SSL-benchmark

Adapting Visual-Language Models for Generalizable Anomaly Detection in Medical Images

中文：适应视觉语言模型以在医学图像中实现泛化的异常检测
Paper: https://arxiv.org/abs/2403.12570
Code: https://github.com/MediaBrain-SJTU/MVFA-AD

FocusMAE: Gallbladder Cancer Detection from Ultrasound Videos with Focused Masked Autoencoders

中文：FocusMAE：聚焦掩蔽自编码器从超声视频中检测胆囊癌
Paper: https://arxiv.org/abs/2403.08848
Code: https://github.com/sbasu276/FocusMAE

Federated Learning（联邦学习）

Think Twice Before Selection: Federated Evidential Active Learning for Medical Image Analysis with Domain Shifts

中文：选择前请三思：带有领域转移的医学图像分析的联邦证据主动学习
Paper: https://arxiv.org/abs/2312.02567

Medical Pre-training $ Foundation Model（预训练&基础模型）

VoCo: A Simple-yet-Effective Volume Contrastive Learning Framework for 3D Medical Image Analysis

中文：VoCo：一个简单而有效的三维医学图像分析的体对比学习框架
Paper: https://arxiv.org/abs/2402.17300
Code: https://github.com/Luffy03/VoCo

MLIP: Enhancing Medical Visual Representation with Divergence Encoder and Knowledge-guided Contrastive Learning

中文：MLIP：使用发散编码器和知识引导的对比学习增强医学视觉表示
Paper: https://arxiv.org/abs/2402.02045

[Highlight!] Continual Self-supervised Learning: Towards Universal Multi-modal Medical Data Representation Learning

中文：持续自我监督学习：走向通用的多模态医学数据表示学习
Paper:https://arxiv.org/abs/2311.17597
Code: https://github.com/yeerwen/MedCoSS

Bootstrapping Chest CT Image Understanding by Distilling Knowledge from X-ray Expert Models

中文：从X射线专家模型中提炼知识以启动胸部CT图像理解
Paper: https://arxiv.org/abs/2404.04936v1

Unleashing the Potential of SAM for Medical Adaptation via Hierarchical Decoding

Paper: https://arxiv.org/abs/2403.18271
Code: https://github.com/Cccccczh404/H-SAM

Low-Rank Knowledge Decomposition for Medical Foundation Models

中文：通过层次解码释放SAM在医学适应中的潜力
Paper: https://arxiv.org/abs/2404.17184
Code: https://github.com/MediaBrain-SJTU/LoRKD

Vision-Language Model (视觉-语言)

PairAug: What Can Augmented Image-Text Pairs Do for Radiology?

中文：PairAug：增强的图像文本对对放射学能做什么？
Paper: https://arxiv.org/abs/2404.04960
Code: https://github.com/YtongXie/PairAug

Decomposing Disease Descriptions for Enhanced Pathology Detection: A Multi-Aspect Vision-Language Matching Framework

中文：疾病描述分解以增强病理检测：一个多方面视觉语言匹配框架
Paper: https://arxiv.org/abs/2403.07636
Code: https://github.com/HieuPhan33/MAVL

Adapting Visual-Language Models for Generalizable Anomaly Detection in Medical Images

中文：适应视觉语言模型以在医学图像中实现泛化的异常检测
Paper: https://arxiv.org/abs/2403.12570
Code: https://github.com/MediaBrain-SJTU/MVFA-AD

OmniMedVQA: A New Large-Scale Comprehensive Evaluation Benchmark for Medical LVLM

中文：OmniMedVQA：一个新的大规模全面评估基准，针对医学LVLM
Paper: https://arxiv.org/abs/2402.09181

CARZero: Cross-Attention Alignment for Radiology Zero-Shot Classification

中文：CARZero：放射学零样本分类的交叉注意力对齐
Paper: https://arxiv.org/abs/2402.17417

FairCLIP: Harnessing Fairness in Vision-Language Learning.

中文：FairCLIP：在视觉语言学习中利用公平性
Paper: https://arxiv.org/abs/2403.19949
Code: https://github.com/Harvard-Ophthalmology-AI-Lab/FairCLIP

Computational Pathology (计算病理)

Generalizable Whole Slide Image Classification with Fine-Grained Visual-Semantic Interaction

中文：具有精细视觉语义交互的可泛化全片图像分类
Paper: https://arxiv.org/abs/2402.19326

Feature Re-Embedding: Towards Foundation Model-Level Performance in Computational Pathology

中文：特征再嵌入：朝着计算病理学的基础模型级性能迈进
Paper: https://arxiv.org/abs/2402.17228
Code: https://github.com/DearCaat/RRT-MIL

PrPSeg: Universal Proposition Learning for Panoramic Renal Pathology Segmentation

中文：PrPSeg：全景肾病病理分割的通用命题学习
Paper: https://arxiv.org/abs/2402.19286

ChAda-ViT: Channel Adaptive Attention for Joint Representation Learning of Heterogeneous Microscopy Images

中文：ChAda-ViT：通道自适应注意力用于异质显微图像的联合表示学习
Paper: https://arxiv.org/abs/2311.15264
Code: https://github.com/nicoboou/chada_vit

SI-MIL: Taming Deep MIL for Self-Interpretability in Gigapixel Histopathology

中文：SI-MIL：驯服深度MIL以实现千兆像素组织病理学的自我解释性
Paper: https://arxiv.org/abs/2312.15010

Transcriptomics-guided Slide Representation Learning in Computational Pathology.

中文：计算病理学中转录组学指导的切片表示学习
Paper: https://arxiv.org/abs/2405.11618
Code: https://github.com/mahmoodlab/TANGLE

Others

Seeing Unseen: Discover Novel Biomedical Concepts via Geometry-Constrained Probabilistic Modeling

中文：看见未见：通过几何约束的概率建模发现新型生物医学概念
Paper: https://arxiv.org/html/2403.01053v2

      
       
      
      
       
                  
           
                         
              
               
                
                 
                  往期精彩回顾
                 
                
               
              
             
             
              
               
                
                 

                
                
                 
                  

                 
                 
                  

                 
                
                
                 

                
               
              
              
               
                
                 适合初学者入门人工智能的路线及资料下载
                
               
                
                 (图文+视频)机器学习入门系列下载
                
               
                
                 机器学习及深度学习笔记等资料打印
                
               
                
                 《统计学习方法》的代码复现专辑

```
交流群
```

欢迎加入机器学习爱好者微信群一起和同行交流，目前有机器学习交流群、博士群、博士申报交流、CV、NLP等微信群，请扫描下面的微信号加群，备注：”昵称-学校/公司-研究方向“，例如：”张小明-浙大-CV“。请按照格式备注，否则不予通过。添加成功后会根据研究方向邀请进入相关微信群。请勿在群内发送广告，否则会请出群，谢谢理解~（也可以加入机器学习交流qq群772479961）