MICCAI 2024 Papers — Page 8
International Conference on Medical Image Computing and Computer-Assisted Intervention · 856 papers
SimTxtSeg: Weakly-Supervised Medical Image Segmentation with Simple Text Cues
Xie, Yuxin (Southeast University), Chen, Geng (Northwestern Polytechnical University)
SegmentationConvolutional Neural NetworkLarge Language ModelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Using simple text prompts to achieve weakly supervised medical image segmentation, the SimTxtSeg framework is constructed, which first generates visual prompts from text to obtain pseudo-labels, and then trains the segmentation model.
Simulation Based Inference for PET iterative reconstruction
Bergere, Bastien (CEA, LIST), Comtat, Claude (Université Paris-Saclay)
Mixture of ExpertsImageBiomedical DataMagnetic Resonance ImagingPositron Emission Tomography
🎯 What it does: A neural density estimator is learned to approximate the PET system matrix using continuous probability distributions generated by a Monte Carlo simulator, enabling simulation-based iterative reconstruction.
Simulation-Based Segmentation of Blood Vessels in Cerebral 3D OCTA Images
Wittmann, Bastian (University of Zurich), Menze, Bjoern (University of Zurich)
SegmentationData SynthesisConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: A method for unsupervised segmentation of intra-brain 3D OCTA vessels based on synthetic data is proposed.
Simultaneous Monocular Endoscopic Dense Depth and Odometry Estimation Using Local-Global Integration Networks
Fan, Wenkang (Xiamen University), Luo, Xiongbiao (Xiamen University)
Pose EstimationDepth EstimationConvolutional Neural NetworkTransformerSimultaneous Localization and MappingVideo
🎯 What it does: A network for unsupervised monocular endoscopic dense depth and pose estimation called LGIN is proposed, based on a hybrid encoder of CNN and Pyramid Vision Transformer (PVT).
Simultaneous Tri-Modal Medical Image Fusion and Super-Resolution using Conditional Diffusion Model
Xu, Yushen (Foshan University), Tan, Haishu (Foshan University)
Image TranslationSuper ResolutionDiffusion modelImageMultimodalityBiomedical DataMagnetic Resonance ImagingComputed TomographyPositron Emission Tomography
🎯 What it does: An end-to-end model TFS-Diff for synchronous fusion and super-resolution of tri-modal medical images has been implemented.
Single-source Domain Generalization in Deep Learning Segmentation via Lipschitz Regularization
Arslan, Mazlum Ferhat (Middle East Technical University), Li, Shuo (Harbin Institute of Technology)
SegmentationDomain AdaptationConvolutional Neural NetworkContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper addresses the single-source domain generalization problem and proposes a spectrum-based Lipschitz regularization method (LRFS) for medical image segmentation.
SiNGR: Brain Tumor Segmentation via Signed Normalized Geodesic Transform Regression
Dang, Trung (University of Oulu), Tiulpin, Aleksei (University of Oulu)
SegmentationConvolutional Neural NetworkTransformerImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Proposes brain tumor segmentation through Signed Normalized Geodesic Transform Regression (SiNGR) using soft label regression.
SinoSynth: A Physics-based Domain Randomization Approach for Generalizable CBCT Image Enhancement
Pang, Yunkui (University of North Carolina at Chapel Hill), Lian, Jun (University of North Carolina at Chapel Hill)
RestorationData SynthesisGenerative Adversarial NetworkImageBiomedical DataComputed Tomography
🎯 What it does: A physics-based domain randomization method called SinoSynth is proposed, which can synthesize various CBCT artifacts on CT images and automatically generate aligned CBCT-CT paired training data.
SIX-Net: Spatial-context Information miX-up for Electrode Landmark Detection
Wang, Xinyi (University of Science and Technology of China), Zhou, S. Kevin (Hohai University)
Object DetectionSegmentationConvolutional Neural NetworkImage
🎯 What it does: This paper proposes SIX-Net, a network that achieves precise localization of electrode markings in ray images by integrating mixed spatial context information.
SkinCON: Towards consensus for the uncertainty of skin cancer sub-typing through distribution regularized adaptive predictive sets (DRAPS)
Ren, Zhihang (University of California Berkeley), Whitney, David (University of California Berkeley)
ClassificationConvolutional Neural NetworkImage
🎯 What it does: Proposed the SkinCON dataset and the DRAPS method for uncertainty quantification and prediction set generation in skin cancer subtype diagnosis.
Slice-Consistent 3D Volumetric Brain CT-to-MRI Translation with 2D Brownian Bridge Diffusion Model
Choo, Kyobin (Yonsei University), Hwang, Seong Jae (Mediwhale)
Image TranslationGenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: Utilizing a 2D Brownian Bridge diffusion model to achieve high-quality 3D slice consistency translation from CT to MRI.
Slice-Consistent Lymph Nodes Detection Transformer in CT Scans via Cross-slice Query Contrastive Learning
Yu, Qinji (Shanghai Jiao Tong University), Jin, Dakai (Alibaba Group)
Object DetectionTransformerContrastive LearningImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a detection Transformer based on 2.5D feature fusion and introduces cross-slice query contrastive learning to enhance the slice consistency of lymph node detection.
SlicerTMS: Real-Time Visualization of Transcranial Magnetic Stimulation for Mental Health Treatment
Franke, Loraine (University of Massachusetts Boston), Haehn, Daniel (University of Massachusetts Boston)
Convolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingDiffusion Tensor Imaging
🎯 What it does: We propose and implement SlicerTMS, an open-source system capable of rapidly predicting and visualizing transcranial magnetic stimulation electric fields in a real-time brain navigation environment using deep learning.
SlideGCD: Slide-based Graph Collaborative Training with Knowledge Distillation for Whole Slide Image Classification
Shu, Tong (Hefei University of Technology), Zheng, Yushan (Beihang University)
ClassificationKnowledge DistillationGraph Neural NetworkImageBiomedical Data
🎯 What it does: The SlideGCD framework is proposed, treating the entire pathological slide as a graph node, constructing a large-scale slide-level graph, and enhancing WSI classification performance through knowledge distillation.
SOM2LM: Self-Organized Multi-Modal Longitudinal Maps
Ouyang, Jiahong, Pohl, Kilian M. (Stanford)
Anomaly DetectionAuto EncoderMultimodalityBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease
🎯 What it does: A self-supervised multimodal longitudinal mapping model SOM2LM is designed, which maps each imaging modality to a two-dimensional self-organizing map to capture the disease progression trajectory.
Sparse Bayesian Networks: Efficient Uncertainty Quantification in Medical Image Analysis
Abboud, Zeinab (Polytechnique Montreal), Kadoury, Samuel (CHUM Hospital Research Center)
ClassificationSegmentationImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a sparse Bayesian network training process that can efficiently quantify predictive uncertainty in medical image classification and segmentation tasks.
Sparsity- and Hybridity-Inspired Visual Parameter-Efficient Fine-Tuning for Medical Diagnosis
Liu, Mingyuan (Beihang University), Zhang, Jicong (Beihang University)
ClassificationSegmentationTransformerSupervised Fine-TuningImageBiomedical DataComputed TomographyUltrasound
🎯 What it does: A novel parameter-efficient fine-tuning method based on sparsity and mixture, SH-PEFT, is proposed to achieve the transfer of large visual models to medical diagnostic tasks using a small number of key weights.
Spatial Context Awareness in Surgery through Sound Source Localization
Seibold, Matthias (Research in Orthopedic Computer Science, Balgrist University Hospital), Fürnstahl, Philipp (University of Zurich)
Object DetectionPose EstimationMultimodalityAudio
🎯 What it does: Introducing sound source localization (SSL) technology in the operating room, utilizing acoustic cameras to generate sound intensity heat maps and aligning them with visual images, verifying its feasibility in two tasks: target detection of surgical sawing and keypoint detection of chiseling.
Spatial Diffusion for Cell Layout Generation
Li, Chen (Stony Brook University), Chen, Chao (Harvard Medical School)
Object DetectionGenerationData SynthesisDiffusion modelImage
🎯 What it does: A spatial distribution-guided diffusion model is proposed for generating cell layouts and pathological images, thereby enhancing cell detection performance.
Spatial Transcriptomics Analysis of Zero-shot Gene Expression Prediction
Yang, Yan (Australian National University), Stone, Eric (Australian National University)
Graph Neural NetworkTransformerLarge Language ModelPrompt EngineeringBiomedical Data
🎯 What it does: Proposes a Semantic Guided Network (SGN) to achieve zero-shot gene expression prediction based on spatial transcriptomics image windows.
Spatial-aware Attention Generative Adversarial Network for Semi-supervised Anomaly Detection in Medical Image
Zhang, Zerui (Wuhan University), Xu, Yongchao (Wuhan University)
Anomaly DetectionGenerative Adversarial NetworkImageBiomedical Data
🎯 What it does: A spatial attention-based Generative Adversarial Network (SAGAN) is proposed, which generates healthy images to restore abnormal regions and uses image differences as anomaly scores for semi-supervised medical image anomaly detection.
Spatial-Division Augmented Occupancy Field for Bone Shape Reconstruction from Biplanar X-Rays
Chen, Jixiang (Hong Kong University of Science and Technology), Li, Xiaomeng (Southern Medical University)
SegmentationGenerationKnowledge DistillationImageMeshBenchmark
🎯 What it does: A spatially partitioned enhanced occupancy field (SdAOF) framework is proposed for reconstructing the 3D shape of bones from dual-plane X-ray images.
Spatio-temporal Contrast Network for Data-efficient Learning of Coronary Artery Disease in Coronary CT Angiography
Ma, Xinghua (Harbin Institute of Technology), Li, Shuo (Harbin Institute of Technology)
ClassificationData-Centric LearningConvolutional Neural NetworkTransformerContrastive LearningBiomedical DataComputed Tomography
🎯 What it does: A SC-Net model aimed at the scarcity of coronary CT angiography data is proposed, achieving efficient automatic diagnosis of coronary artery diseases.
Spatio-temporal neural distance fields for conditional generative modeling of the heart
Sørensen, Kristine (DTU Compute Technical University Of Denmark), Paulsen, Rasmus (DTU Compute Technical University Of Denmark)
GenerationData SynthesisAuto EncoderImageBiomedical DataComputed Tomography
🎯 What it does: A conditional generative model based on neural implicit distance fields (SDF) is proposed, which can complete the entire cardiac cycle given static atrial images and clinical demographic information, and generate new dynamic atrial sequences based on clinical information.
Spatiotemporal Graph Neural Network Modelling Perfusion MRI
Yan, Ruodan (University of Cambridge), Li, Chao (University of Dundee)
Graph Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A spatiotemporal graph neural network called PerfGAT is proposed to predict the IDH mutation status of brain gliomas using dynamic contrast-enhanced MRI.
Spatiotemporal Representation Learning for Short and Long Medical Image Time Series
Shen, Chengzhi (Technical University of Munich), Holland, Robbie (Imperial College London)
Representation LearningTransformerContrastive LearningVideoTime SeriesBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes two self-supervised spatiotemporal representation learning methods, namely c SimCLR based on single-slice contrastive learning and TVRL which combines frame-level feature prediction, to improve the representation of medical image time series.
SpeChrOmics: A Biomarker Characterization Framework for Medical Hyperspectral Imaging
Oladokun, Ajibola S. (University of Cape Town), Mutsvangwa, Tinashe E. M. (University of Cape Town)
ClassificationExplainability and InterpretabilityImageBiomedical Data
🎯 What it does: Proposes the SpeChrOmics framework, which integrates spectral unmixing and spatial radiomics to automatically extract potential biomarkers from medical hyperspectral images;
Spot the Difference: Difference Visual Question Answering with Residual Alignment
Lu, Zilin (Northwestern Polytechnical University), Xia, Yong (Northwestern Polytechnical University)
RecognitionGenerationConvolutional Neural NetworkTransformerLarge Language ModelContrastive LearningImageTextMultimodalityElectronic Health Records
🎯 What it does: A residual alignment model called ReAl is proposed for differential visual question answering (DiffVQA), which generates answers about differences by comparing chest X-ray images of the same patient at different time points.
SRE-CNN: A Spatiotemporal Rotation-Equivariant CNN for Cardiac Cine MR Imaging
Zhu, Yuliang (University of Nottingham Ningbo China), Liang, Dong (Chinese Academy of Sciences)
RestorationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a spatiotemporal rotation equivariant convolutional neural network (SRE-CNN) for high-acceleration reconstruction of dynamic cardiac CINE MRI.
Stable Diffusion Segmentation for Biomedical Images with Single-step Reverse Process
Lin, Tianyu (Sun Yat-sen University), Zheng, Fudan (Sun Yat-sen University)
SegmentationDiffusion modelImageBiomedical DataComputed TomographyUltrasound
🎯 What it does: This paper presents SDSeg, a medical image segmentation framework based on Stable Diffusion, which can generate segmentation results in a single-step reverse process, eliminating the need for traditional multi-step sampling and multiple sample averaging.
STAN-LOC: Visual Query-based Video Clip Localization for Fetal Ultrasound Sweep Videos
Mishra, Divyanshu (University of Oxford), Noble, J. Alison (University of Oxford)
RetrievalTransformerContrastive LearningVideoUltrasound
🎯 What it does: This paper proposes and implements a visual query video clipping localization task (VQ-VCL) for fetal ultrasound sweep videos, aimed at automatically retrieving segments containing standard anatomical views.
Stealing Knowledge from Pre-trained Language Models for Federated Classifier Debiasing
Zhu, Meilu (City University of Hong Kong), Yuan, Yixuan (Southern University of Science and Technology)
ClassificationFederated LearningConvolutional Neural NetworkLarge Language ModelPrompt EngineeringImageTextBiomedical Data
🎯 What it does: The FedCB framework is proposed, which constructs a fixed classifier by extracting text embeddings from pre-trained language models and aligns image features with text distributions during the federated learning process, thereby addressing the classifier bias problem caused by heterogeneous data.
StereoDiffusion: Temporally Consistent Stereo Depth Estimation with Diffusion Models
Xu, Haozheng (Imperial College London), Giannarou, Stamatia (Imperial College London)
Depth EstimationDiffusion modelOptical FlowVideoBiomedical Data
🎯 What it does: Proposes the StereoDiffusion framework, which utilizes latent diffusion models and optical flow priors to achieve temporally consistent disparity estimation in medical surgical videos.
Stochastic Anomaly Simulation for Maxilla Completion from Cone-Beam Computed Tomography
Guo, Yixiao (Peking University), Zha, Hongbin (Peking University)
RestorationSegmentationData SynthesisAnomaly DetectionConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a Stochastic Anomaly Simulation (SAS) algorithm to generate defective CBCT data, and implements jawbone defect recovery and defect mask prediction based on a weakly supervised voxel-level image restoration framework.
Striving for Simplicity: Simple Yet Effective Prior-Aware Pseudo-Labeling for Semi-Supervised Ultrasound Image Segmentation
Chen, Yaxiong (Wuhan University of Technology), Mou, Lichao (Xidian University)
SegmentationConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical DataUltrasound
🎯 What it does: This paper proposes a semi-supervised ultrasound image segmentation framework based on pseudo-labels, and constrains the segmentation results through adversarial learning of shape priors, addressing the issue of meaningless segmentation that can occur in the absence of labels.
Structural Attention: Rethinking Transformer for Unpaired Medical Image Synthesis
Phan, Vu Minh Hieu (University of Adelaide), To, Minh-Son (University of Adelaide)
GenerationData SynthesisTransformerGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance ImagingComputed TomographyPositron Emission Tomography
🎯 What it does: This paper proposes a structure attention network UNest based on Transformer for unpaired medical image synthesis.
Structural Entities Extraction and Patient Indications Incorporation for Chest X-ray Report Generation
Liu, Kang (Xidian University), Miao, Qiguang (Brown University)
GenerationTransformerImageTextMultimodalityElectronic Health RecordsRetrieval-Augmented Generation
🎯 What it does: A chest X-ray report generation framework (SEI) that combines structural entity extraction and patient indication fusion is proposed.
Structure-preserving Image Translation for Depth Estimation in Colonoscopy
Wang, Shuxian (University of North Carolina at Chapel Hill), Sengupta, Roni (University of North Carolina at Chapel Hill)
Image TranslationDepth EstimationConvolutional Neural NetworkGenerative Adversarial NetworkImage
🎯 What it does: A structure-preserving synthetic-to-real image translation pipeline is proposed to convert synthetic colonoscopy images into realistic appearances while preserving depth geometry, aiming to enhance monocular depth estimation performance.
Subgroup-Specific Risk-Controlled Dose Estimation in Radiotherapy
Fischer, Paul (University of Tübingen), Baumgartner, Christian F. (University of Tübingen)
Convolutional Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: In radiation therapy, the subgroup risk control prediction set (SG-RCPS) provides a reliable uncertainty interval for deep learning dose prediction.
Subject-Adaptive Transfer Learning Using Resting State EEG Signals for Cross-Subject EEG Motor Imagery Classification
An, Sion (Daegu Gyeongbuk Institute of Science and Technology), Park, Sang Hyun (DGIST)
ClassificationDomain AdaptationConvolutional Neural NetworkTransformerTime SeriesBiomedical Data
🎯 What it does: An adaptive transfer learning framework called ResTL is proposed, based on resting-state EEG signals, for motor imagery (MI) classification in cross-subject brain-computer interfaces.
Super-Field MRI Synthesis for Infant Brains Enhanced by Dual Channel Latent Diffusion
Tapp, Austin (Children's National Hospital), Linguraru, Marius George (George Washington University)
SegmentationGenerationData SynthesisDiffusion modelGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A network named SFNet has been developed, utilizing ultra-low field (0.064T) infant MRI to generate dual-channel paired data through latent diffusion, producing high-quality images equivalent to high-field (≥1.5T) MRI;
Superpixel-Guided Segment Anything Model for Liver Tumor Segmentation with Couinaud Segment Prompt
Lyu, Fei (Hong Kong Baptist University), Yuen, Pong C. (Chinese University of Hong Kong)
SegmentationTransformerPrompt EngineeringBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a novel framework called CouinaudSAM, which combines Couinaud segment text prompts with the Segment Anything Model (SAM) for liver tumor segmentation.
Surface-based and Shape-informed U-fiber Atlasing for Robust Superficial White Matter Connectivity Analysis
Li, Yuan (University of Southern California), Shi, Yonggang (University of Southern California)
Biomedical DataMagnetic Resonance ImagingDiffusion Tensor ImagingAlzheimer's Disease
🎯 What it does: A surface substrate dictionary containing 77 main U-fiber bundles was constructed, and a personalized U-fiber imaging method based on shape similarity was proposed to reconstruct complete SWM connectivity on low-resolution clinical dMRI.
Surgformer: Surgical Transformer with Hierarchical Temporal Attention for Surgical Phase Recognition
Yang, Shu (Hong Kong University of Science and Technology), Chen, Hao (Harvard Medical School)
RecognitionTransformerVideo
🎯 What it does: This paper proposes an end-to-end Surgformer model that achieves surgical phase recognition through sparse frame sequences using separated spatiotemporal attention.
SurgicalGaussian: Deformable 3D Gaussians for High-Fidelity Surgical Scene Reconstruction
Xie, Weixing (Xiamen University), Guo, Xiaohu (University of Texas at Dallas)
RestorationGenerationDepth EstimationGaussian SplattingPoint Cloud
🎯 What it does: This paper proposes SurgicalGaussian, which implements high-fidelity dynamic scene reconstruction for endoscopy based on a deformable 3D Gaussian scattering model.
Survival analysis of histopathological image based on a pretrained hypergraph model of spatial transcriptomics data
Cai, Shangyan (Beijing Normal University-Hong Kong Baptist University United International College), Su, Weifeng (Beijing Normal-Hong Kong Baptist University)
Graph Neural NetworkImageMultimodalityBiomedical Data
🎯 What it does: This paper proposes a multimodal hypergraph neural network (MHNN-surv) that integrates histopathological images with spatial gene expression information through a pretrained spatial transcriptomics model to perform survival analysis for breast cancer.
SurvRNC: Learning Ordered Representations for Survival Prediction using Rank-N-Contrast
Saeed, Numan (Mohamed bin Zayed University of Artificial Intelligence), Yaqub, Mohammad (Mohamed bin Zayed University of Artificial Intelligence)
Representation LearningConvolutional Neural NetworkContrastive LearningMultimodalityBiomedical DataComputed TomographyPositron Emission TomographyElectronic Health Records
🎯 What it does: A new SurvRNC loss function is proposed and implemented on multimodal head and neck cancer data to learn latent representations ordered by survival time, and it is integrated into deep survival models (DeepMTLR, DeepHit).
Swin SMT: Global Sequential Modeling for Enhancing 3D Medical Image Segmentation
Płotka, Szymon (University of Amsterdam), Biecek, Przemyslaw (Warsaw University of Technology)
SegmentationTransformerMixture of ExpertsImageBiomedical DataComputed Tomography
🎯 What it does: A Swin SMT model based on Swin Transformer is proposed, enhancing global sequence modeling with Soft MoE to achieve 3D segmentation of 117 anatomical structures in whole-body CT (WBCT) images.
Swin-UMamba: Mamba-based UNet with ImageNet-based pretraining
Liu, Jiarun (Paul C. Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences), Wang, Shanshan (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences)
SegmentationConvolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A Mamba-based UNet network called Swin-UMamba is proposed, specifically for medical image segmentation, and its performance is enhanced through ImageNet pre-training.
Symmetry Awareness Encoded Deep Learning Framework for Brain Imaging Analysis
Ma, Yang (University of Sydney), Wang, Chenyu (University of Sydney)
ClassificationSegmentationTransformerContrastive LearningImageMultimodalityBiomedical DataMagnetic Resonance ImagingComputed TomographyAlzheimer's Disease
🎯 What it does: A Symmetry-Aware Cross-Attention (SACA) module and Symmetry-Aware Head (SAH) based on the symmetrical features of the left and right hemispheres are proposed, enhancing brain imaging classification and segmentation performance through symmetry pre-training.
Symptom Disentanglement in Chest X-ray Images for Fine-Grained Progression Learning
Zhu, Ye (Hong Kong Baptist University), Yuen, Pong C. (Chinese University of Hong Kong)
ClassificationRecognitionTransformerImageBiomedical Data
🎯 What it does: A chest X-ray progression learning framework based on symptom disentanglement is proposed, utilizing a Symptom Disentangler to extract symptom features and capturing symptom-level progression through a Symptom Progression Learner.
SynCellFactory: Generative Data Augmentation for Cell Tracking
Sturm, Moritz (Heidelberg University), Hamprecht, Fred A. (Heidelberg University)
Object TrackingGenerationData SynthesisDiffusion modelVideoBiomedical Data
🎯 What it does: Construct a generative data augmentation pipeline called SynCellFactory based on ControlNet, designed for the automatic generation of cell videos with realistic segmentation and tracking labels.
Synchronous Image-Label Diffusion with Anisotropic Noise for Stroke Lesion Segmentation on Non-contrast CT
Zhang, Jianhai, Ganesh, Aravind (University Of Calgary)
SegmentationConvolutional Neural NetworkDiffusion modelImageComputed Tomography
🎯 What it does: This paper proposes a synchronous image-label diffusion model, replacing traditional isotropic Gaussian noise with anisotropic noise for the automatic segmentation of stroke lesions in non-contrast CT images.
TabMixer: Noninvasive Estimation of the Mean Pulmonary Artery Pressure via Imaging and Tabular Data Mixing
Grzeszczyk, Michal K. (Sano Centre for Computational Medicine), Sitek, Arkadiusz (Massachusetts General Hospital)
Convolutional Neural NetworkTransformerVideoTabularBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Using cardiac magnetic resonance (CMR) videos and tabular features, we propose the TabMixer module to construct a deep learning framework for non-invasive measurement of mean pulmonary artery pressure (mPAP).
Tackling Data Heterogeneity in Federated Learning via Loss Decomposition
Zeng, Shuang (The University of Hong Kong), Qu, Liangqiong (The University of Hong Kong)
Federated LearningImageBiomedical Data
🎯 What it does: This paper proposes to decompose the global loss of federated learning into local loss, distribution shift loss, and aggregation loss, and based on this, designs two methods: Margin control regularization and main gradient aggregation, to jointly reduce the three types of losses and achieve more robust federated learning.
TADM: Temporally-Aware Diffusion Model for Neurodegenerative Progression on Brain MRI
Litrico, Mattia (University of Catania), Battiato, Sebastiano (University of Catania)
GenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A method for generating temporal-aware brain MRI progress based on diffusion models is proposed and implemented, predicting future MRIs by learning the intensity differences between baseline and follow-up scans.
TaGAT: Topology-Aware Graph Attention Network For Multi-modal Retinal Image Fusion
Tian, Xin (University of Bristol), Achim, Alin (University of Bristol)
Image TranslationRestorationData SynthesisGraph Neural NetworkTransformerImageMultimodalityBiomedical Data
🎯 What it does: A multi-modal retinal image fusion framework called TaGAT is designed and implemented, which combines spatial features with vascular topological structures through a topology-aware graph attention network to achieve high-quality fused image generation.
Tagged-to-Cine MRI Sequence Synthesis via Light Spatial-Temporal Transformer
Liu, Xiaofeng (Massachusetts General Hospital and Harvard Medical School), Woo, Jonghye (Massachusetts General Hospital and Harvard Medical School)
GenerationData SynthesisTransformerImageVideoBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A method is proposed to synthesize Tagged MRI to Cine MRI sequences using a lightweight spatiotemporal Transformer (LiST 2) and Recursive Sliding Tuning (ReST).
Tail-Enhanced Representation Learning for Surgical Triplet Recognition
Gui, Shuangchun (Southern University of Science and Technology), Wang, Zhenkun (Southern University of Science and Technology)
RecognitionRepresentation LearningTransformerContrastive LearningVideo
🎯 What it does: A tail class enhancement representation learning framework (TERL) based on multi-task learning, instance-level contrastive learning, and prototype semantic enhancement is designed to improve the recognition performance of tail classes in surgical video triplet recognition.
TAKT: Target-Aware Knowledge Transfer for Whole Slide Image Classification
Xiong, Conghao (Chinese University of Hong Kong), King, Irwin (Chinese University of Hong Kong)
ClassificationDomain AdaptationKnowledge DistillationSupervised Fine-TuningImageBiomedical Data
🎯 What it does: This paper proposes a target-aware knowledge transfer framework (TAKT) for whole-slide image classification, achieving knowledge transfer through a teacher-student paradigm that combines target-aware data augmentation and feature alignment.
TAPoseNet: Teeth Alignment based on Pose estimation via multi-scale Graph Convolutional Network
Deng, Qingxin (Shenzhen University), Zhang, Dian (Shenzhen University)
Pose EstimationGraph Neural NetworkAuto EncoderPoint Cloud
🎯 What it does: A deep learning framework named TAPoseNet is proposed for automatically predicting the alignment targets of teeth after orthodontic treatment, avoiding deformation of tooth shapes; it first estimates the pose of each tooth and then uses a multi-scale graph convolutional network to predict the target pose, resulting in the final dental arch model.
TARDRL: Task-Aware Reconstruction for Dynamic Representation Learning of fMRI
Zhao, Yunxi (Nanjing University of Aeronautics and Astronautics), Wen, Xuyun (Nanjing University of Aeronautics and Astronautics)
ClassificationRepresentation LearningTransformerBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: This study proposes a task-aware reconstruction dynamic representation learning framework (TARDRL) for achieving better disease diagnosis prediction in functional magnetic resonance imaging (fMRI).
TE-SSL: Time and Event-aware Self Supervised Learning for Alzheimer’s Disease Progression Analysis
Thrasher, Jacob (West Virginia University), the Alzheimer’s Disease Neuroimaging Initiative
ClassificationRepresentation LearningConvolutional Neural NetworkContrastive LearningImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: A self-supervised learning framework TE-SSL is proposed, which combines event labels and time difference weights for imaging feature extraction and survival analysis of Alzheimer's disease progression.
TeethDreamer: 3D Teeth Reconstruction from Five Intra-oral Photographs
Xu, Chenfan (ShanghaiTech University), Cui, Zhiming (ShanghaiTech University)
GenerationData SynthesisConvolutional Neural NetworkDiffusion modelImage
🎯 What it does: The TeethDreamer framework is proposed, which generates multi-view color images and normal maps from five intraoral photos, and obtains high-quality 3D tooth models through neural surface reconstruction.
TeleOR: Real-time Telemedicine System for Full-Scene Operating Room
Wu, Yixuan (Zhejiang University), Wu, Jian (Zhejiang University)
Pose EstimationDepth EstimationCompressionComputational EfficiencySimultaneous Localization and MappingOptical FlowVideoPoint Cloud
🎯 What it does: This paper implements a real-time panoramic operating room scene reconstruction and transmission system for remote surgical guidance.
Temporal Neighboring Multi-Modal Transformer with Missingness-Aware Prompt for Hepatocellular Carcinoma Prediction
Xu, Jingwen (Hong Kong Baptist University), Yuen, Pong C. (Chinese University of Hong Kong)
TransformerPrompt EngineeringMultimodalityTime SeriesBiomedical DataComputed TomographyElectronic Health Records
🎯 What it does: This paper proposes a Temporal Neighboring Multi-Modal Transformer with Missingness-Aware Prompt (TNformer-MP) for early prediction of hepatocellular carcinoma by integrating clinical time series and CT scans.
Textmatch: Using Text Prompts to Improve Semi-supervised Medical Image Segmentation
Li, Aibing (Sichuan University), Wang, Yan (East China Normal University)
SegmentationConvolutional Neural NetworkTransformerPrompt EngineeringContrastive LearningImageMultimodalityBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a framework called Textmatch that enhances semi-supervised medical image segmentation using text prompts, combining visual and linguistic feature fusion, view consistency regularization, and pseudo-label contrastive learning, significantly improving segmentation performance.
TextPolyp: Point-supervised Polyp Segmentation with Text Cues
Zhao, Yiming (Nanjing University of Science and Technology), Zhou, Tao (Nanjing University of Science and Technology)
SegmentationImage
🎯 What it does: A weakly supervised polyp segmentation method based on SAM and text prompts has been developed, which can generate high-quality segmentation results with only point annotations.
Textual Inversion and Self-supervised Refinement for Radiology Report Generation
Luo, Yuanjiang (Sichuan University), Zhang, Yi (Sichuan University)
GenerationContrastive LearningImageTextBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A TISR method is proposed, which utilizes text inversion to map image features to pseudo-words, and then enhances the accuracy of radiology report generation through self-supervised refinement.
The Centerline-Cross Entropy Loss for Vessel-Like Structure Segmentation: Better Topology Consistency Without Sacrificing Accuracy
Acebes, Cesar (Universitat Pompeu Fabra), Galdran, Adrian (Universitat Pompeu Fabra)
SegmentationConvolutional Neural NetworkBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a new centerline cross-entropy (clCE) loss function for the segmentation of vascular morphology in medical images.
The MRI Scanner as a Diagnostic: Image-less Active Sampling
Du, Yuning (University of Edinburgh), Tsaftaris, Sotirios A. (University of Edinburgh)
OptimizationComputational EfficiencyConvolutional Neural NetworkReinforcement LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A reinforcement learning-based active sampling framework for MRI k-space is proposed, which directly performs lesion diagnosis on un-reconstructed k-space data.
This actually looks like that: Proto-BagNets for local and global interpretability-by-design
Djoumessi, Kerol (University of Tbingen), Koch, Lisa (University of Tbingen)
ClassificationExplainability and InterpretabilityConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an interpretable Proto-BagNet model for detecting drusen, a lesion related to age-related macular degeneration, in retinal optical coherence tomography (OCT) images, providing local and global interpretable prototype images during the inference phase.
ThyGraph: A Graph-Based Approach for Thyroid Nodule Diagnosis from Ultrasound Studies
Radhachandran, Ashwath (UCLA), Speier, William (UCLA)
ClassificationExplainability and InterpretabilityGraph Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: A ThyGraph method based on graph convolutional networks is proposed, which constructs patient-level graphs using multi-view ultrasound images to predict the malignancy of thyroid nodules.
TinyU-Net: Lighter yet Better U-Net with Cascaded Multi-Receptive Fields
Chen, Junren (Sichuan University), Chen, Liangyin (Sichuan University)
SegmentationConvolutional Neural NetworkBiomedical DataComputed Tomography
🎯 What it does: A lightweight TinyU-Net and its core CMRF module are proposed for medical image segmentation.
TLRN: Temporal Latent Residual Networks For Large Deformation Image Registration
Wu, Nian (University of Virginia), Zhang, Miaomiao (University of Virginia)
Convolutional Neural NetworkImageVideoBiomedical DataMagnetic Resonance ImagingOrdinary Differential Equation
🎯 What it does: Design and implement a Temporal Latent Residual Network (TLRN) that predicts continuous deformation fields of time series images by recursively learning the residual function in the latent velocity field space, achieving high-precision time series registration.
Topological Cycle Graph Attention Network for Brain Functional Connectivity
Huang, Jinghan (National University of Singapore), Qiu, Anqi (Hong Kong Polytechnic University)
ClassificationGraph Neural NetworkGraphBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A method based on the Topological Cycle Graph Attention Network (CycGAT) is proposed to identify functional backbones from brain functional connectivity graphs and eliminate redundant edges.
Topological GCN for Improving Detection of Hip Landmarks from B-Mode Ultrasound Images
Huang, Tianxiang (Shanghai University), Shi, Jun (Shanghai University)
Object DetectionPose EstimationConvolutional Neural NetworkGraph Neural NetworkTransformerImageBiomedical DataUltrasound
🎯 What it does: Design and implement a framework that combines an improved Conformer with Topological GCN for hip joint keypoint detection in B-mode ultrasound images.
Topological SLAM in colonoscopies leveraging deep features and topological priors
Morlana, Javier (University of Zaragoza), Montiel, José M. M. (Universidad de Zaragoza)
TransformerSimultaneous Localization and MappingImageVideoBiomedical Data
🎯 What it does: This paper presents ColonSLAM, a system that combines classical multi-map metric SLAM with deep features and topological priors to construct a topological map of the entire colon.
Topologically faithful multi-class segmentation in medical images
Berger, Alexander H. (Technical University of Munich), Paetzold, Johannes C. (Cornell Tech)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A topologically faithful loss function for multi-class medical image segmentation is proposed, which can directly ensure the topological correctness of multi-class segmentation results during the training phase.
Toward Universal Medical Image Registration via Sharpness-Aware Meta-Continual Learning
Wang, Bomin (Fudan University), Zhuang, Xiahai (Fudan University)
Meta LearningConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A Sharpness-Aware Meta-Continual Learning (SAMCL) framework is proposed for achieving general registration of 3D medical images.
Towards a Benchmark for Colorectal Cancer Segmentation in Endorectal Ultrasound Videos: Dataset and Model Development
Jiang, Yuncheng (Shenzhen Future Network of Intelligence Institute), Li, Zhen (SSE Chinese University of Hong Kong)
SegmentationTransformerImageVideoUltrasoundBenchmark
🎯 What it does: This paper proposes a framework for automatic segmentation of colorectal cancer using ERUS videos, constructing the first ERUS-10K dataset containing segmentation and infiltration depth annotations, and developing the ASTR model based on this dataset.
Towards a Deeper insight into Face Detection in Neonatal wards
Zhao, Yisheng (Zhejiang University), Pan, Yun (Zhejiang University)
Object DetectionConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: This paper first constructs the first publicly available neonatal facial detection dataset NFD and proposes an improved detection framework on this dataset.
Towards a text-based quantitative and explainable histopathology image analysis
Nguyen, Anh Tien (Korea University), Kwak, Jin Tae (Korea University)
ClassificationRetrievalExplainability and InterpretabilityTransformerVision Language ModelImageTextMultimodalityBiomedical Data
🎯 What it does: The TQx framework is proposed, utilizing a pre-trained vision-language model to generate interpretable text features through image-text retrieval, achieving quantitative analysis of pathological images.
Towards Explainable Automated Neuroanatomy
Qian, Kui (University of California, San Diego), Freund, Yoav (University of California, San Diego)
ClassificationExplainability and InterpretabilityDiffusion modelBiomedical Data
🎯 What it does: Utilizing cell shape features for interpretable automatic identification of brain structures, constructing region features based on cell distribution and using XGBoost for classification;
Towards Graph Neural Networks with Domain-Generalizable Explainability for fMRI-Based Brain Disorder Diagnosis
Qiu, Xinmei, Ma, Jianhua (Pazhou Lab)
Domain AdaptationExplainability and InterpretabilityMeta LearningGraph Neural NetworkBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a graph neural network called XG-GNN that can simultaneously achieve interpretability and domain generalization for multi-center fMRI brain disease diagnosis.
Towards Integrating Epistemic Uncertainty Estimation into the Radiotherapy Workflow
Teichmann, Marvin Tom (Siemens Healthineers), Ghesu, Florin C. (Siemens Healthineers)
SegmentationAnomaly DetectionConvolutional Neural NetworkBiomedical DataComputed Tomography
🎯 What it does: Integrate uncertainty estimation techniques into the organ segmentation workflow in radiotherapy, and implement detection and alerts for clinically relevant out-of-distribution (OOD) scenarios.
Towards Learning Contrast Kinetics with Multi-Condition Latent Diffusion Models
Osuala, Richard (Universitat de Barcelona), Lekadir, Karim (Institució Catalana de Recerca i Estudis Avançats)
Image TranslationGenerationData SynthesisDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This study investigates how to utilize a multi-condition latent diffusion model (CC-Net) to convert pre-contrast images of the breast into post-contrast images, thereby simulating the contrast agent uptake process in dynamic contrast-enhanced MRI (DCE-MRI).
Towards Multi-modality Fusion and Prototype-based Feature Refinement for Clinically Significant Prostate Cancer Classification in Transrectal Ultrasound
Wu, Hong (Shenzhen University), Wang, Yi (Shenzhen University)
ClassificationConvolutional Neural NetworkVideoMultimodalityUltrasound
🎯 What it does: A deep learning framework based on multimodal transrectal ultrasound (TRUS) video is proposed, which achieves the classification of clinically significant prostate cancer (csPCa) and generates visual activation maps through a fusion of attention mechanisms and few-shot segmentation assistance.
Towards Precise Pose Estimation in Robotic Surgery: Introducing Occlusion-Aware Loss
Park, Jihun (Kyungpook National University), Jung, Heechul (hutom)
Pose EstimationRobotic IntelligenceVideo
🎯 What it does: A novel occlusion-aware loss function for robotic surgical tool pose estimation is proposed, and a private dataset based on real gastric cancer surgical videos is constructed.
Towards Rapid Mycetoma Species Diagnosis: A Deep Learning Approach for Stain-Invariant Classification on H&E Images from Senegal
Zinsou, Kpêtchéhoué Merveille Santi (University of Gaston Berger), Ndiaye, Maodo (University Cheikh Anta Diop)
ClassificationRecognitionConvolutional Neural NetworkImage
🎯 What it does: An automated diagnostic system based on H&E histological images has been developed, utilizing staining standardization and a CNN model for the rapid identification of black skin Mycetoma pathogens.
Towards Real-time Intrahepatic Vessel Identification in Intraoperative Ultrasound-Guided Liver Surgery
Beaudet, Karl-Philippe (IHU Strasbourg), Verde, Juan (IHU Strasbourg)
RecognitionSegmentationConvolutional Neural NetworkImageBiomedical DataUltrasound
🎯 What it does: This paper studies the generation of 3D ultrasound volumes from preoperative 2D ultrasound scans through registration, and trains a personalized Attention U-Net model based on this volume to achieve real-time identification of intrahepatic vessels during laparoscopic surgery.
Towards realistic needle insertion training simulator using partitioned model order reduction
Vanneste, Félix, Duriez, Christian (InfinyTech3D)
Computational EfficiencyBiomedical Data
🎯 What it does: A real-time liver acupuncture training simulator based on a partitioned model for sequential dimensionality reduction is proposed.
Towards tDCS Digital Twins using Deep Learning-based Direct Estimation of Personalized Electrical Field Maps from T1-Weighted MRI
Stolte, Skylar E. (University of Florida), Fang, Ruogu (University of Florida)
Convolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a deep learning-based framework that directly predicts individualized tDCS electric field distribution from T1-weighted MRI, eliminating the need for traditional finite element simulations.
TP-DRSeg: Improving Diabetic Retinopathy Lesion Segmentation with Explicit Text-Prompts Assisted SAM
Li, Wenxue (Monash University), Ge, Zongyuan (Melbourne University)
SegmentationConvolutional Neural NetworkPrompt EngineeringContrastive LearningImageBiomedical Data
🎯 What it does: A text prompt-assisted Segment Anything Model (TP-DRSeg) framework is proposed for the segmentation of diabetic retinopathy (DR) lesions.
TractOracle: towards an anatomically-informed reward function for RL-based tractography
Théberge, Antoine, Jodoin, Pierre-Marc (Universit' de Sherbrooke)
TransformerReinforcement LearningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A reinforcement learning-driven fiber bundle tracking system named TractOracle is proposed, utilizing a Transformer network as an anatomical credibility assessor, which can serve as a reward function during training and can also stop tracking in real-time to generate high-precision fiber bundles.
Training ViT with Limited Data for Alzheimer’s Disease Classification: an Empirical Study
Kunanbayev, Kassymzhomart (KAIST), Kim, Dae-Shik (KAIST)
ClassificationKnowledge DistillationTransformerAuto EncoderImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease
🎯 What it does: Using pure Vision Transformer (ViT) for Alzheimer's classification with limited brain MRI data, this study systematically explores the effects of various training strategies such as self-supervised pre-training, knowledge distillation, and data augmentation.
Training-Free Condition Video Diffusion Models for single frame Spatial-Semantic Echocardiogram Synthesis
Nguyen, Van Phi (Vietnam National University), Tran, Quoc Long (VinUni)
SegmentationGenerationData SynthesisDiffusion modelImageVideoBiomedical DataUltrasoundStochastic Differential Equation
🎯 What it does: A training-free video diffusion model (Free-Echo) is proposed, which can generate realistic cardiac ultrasound videos based solely on a single frame of cardiac ultrasound segmentation images.
Trans-Window Panoramic Impasto for Online Tissue Deformation Recovery
Chen, Jiahe, Tomii, Naoki
RestorationOptimizationNeural Radiance FieldPoint CloudMesh
🎯 What it does: A three-dimensional surface reconstruction framework based on local parameter mapping and gradient consistency constraints is proposed, which solves the correspondence between local mapping and surface through nonlinear least squares optimization, enabling high-precision recovery of complex details.
Transferring Relative Monocular Depth to Surgical Vision with Temporal Consistency
Budd, Charlie (King's College London), Vercauteren, Tom (King's College London)
Depth EstimationDomain AdaptationTransformerOptical FlowImageVideoComputed Tomography
🎯 What it does: Transfer the relative monocular depth model trained on natural images to endoscopic images, significantly improving depth estimation accuracy through time-consistent self-supervision.
Transforming Surgical Interventions with Embodied Intelligence for Ultrasound Robotics
Xu, Huan (Centre for Artificial Intelligence and Robotics, HKISI-CAS), Liu, Hongbin (Centre for Artificial Intelligence and Robotics, Hong Kong Institute of Science Innovation, Chinese Academy of Sciences)
Robotic IntelligenceTransformerLarge Language ModelTextBiomedical DataUltrasoundRetrieval-Augmented Generation
🎯 What it does: This paper proposes an embedded intelligent system based on large language models and domain knowledge enhancement, achieving closed-loop control for precise motion planning and dynamic execution of ultrasound robots from voice commands.