MICCAI 2024 Papers — Page 9
International Conference on Medical Image Computing and Computer-Assisted Intervention · 856 papers
Trexplorer: Recurrent DETR for Topologically Correct Tree Centerline Tracking
Naeem, Roman (Chalmers University of Technology), Kahl, Fredrik (Chalmers University of Technology)
Object TrackingSegmentationTransformerBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Trexplorer is a recursive DETR model designed to track the centerlines of tree structures in 3D medical images, ensuring topological correctness without the need for post-processing.
Tri-modal Confluence with Temporal Dynamics for Scene Graph Generation in Operating Rooms
Guo, Diandian (Chinese University of Hong Kong), Heng, Pheng-Ann (Chinese University of Hong Kong)
Object DetectionGenerationKnowledge DistillationConvolutional Neural NetworkTransformerLarge Language ModelImageVideoTextMultimodalityPoint Cloud
🎯 What it does: An end-to-end tri-modal (image, point cloud, language) temporal dynamic scene graph generation framework called TriTemp-OR is proposed for the operating room scene graph (OR-SGG) task.
Tri-Plane Mamba: Efficiently Adapting Segment Anything Model for 3D Medical Images
Wang, Hualiang (Hong Kong University of Science and Technology), Li, Xiaomeng (Southern Medical University)
SegmentationComputational EfficiencyConvolutional Neural NetworkTransformerSupervised Fine-TuningImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes the TP-Mamba adapter, which migrates the Segment Anything Model (SAM) to 3D medical image segmentation tasks, achieving efficient adaptation in terms of parameters and data.
TrIND: Representing Anatomical Trees by Denoising Diffusion of Implicit Neural Fields
Sinha, Ashish (Simon Fraser University), Hamarneh, Ghassan (Simon Fraser University)
GenerationData SynthesisCompressionTransformerDiffusion modelPoint CloudMesh
🎯 What it does: This paper proposes an anatomical tree representation method called TrIND based on implicit neural representations (INR), and learns the distribution of tree structures through a diffusion model to achieve high-quality, arbitrary-resolution tree reconstruction and generation.
TSBP: Improving Object Detection in Histology Images via Test-time Self-guided Bounding-box Propagation
Yang, Tingting (Nanjing University of Science and Technology), Zhang, Yizhe (Nanjing University of Science and Technology)
Object DetectionConvolutional Neural NetworkDiffusion modelImageBiomedical Data
🎯 What it does: A method called Test-time Self-guided Bounding Box Propagation (TSBP) is proposed, which allows high-confidence detection boxes to influence low-confidence boxes, thereby improving the recall and accuracy of object detection in histology images without the need for additional labeled samples.
Two Projections Suffice for Cerebral Vascular Reconstruction
Cafaro, Alexandre (Harvard Medical School), Frisken, Sarah (Harvard Medical School)
SegmentationOptimizationSupervised Fine-TuningBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A method for reconstructing the three-dimensional structure of cerebral blood vessels from only two DSA projections is proposed.
UinTSeg: Unified Infant Brain Tissue Segmentation with Anatomy Delineation
Liu, Jiameng (ShanghaiTech University), Shen, Dinggang (Shanghai United Imaging Intelligence Co., Ltd.)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a unified infant brain tissue segmentation framework called UinTSeg, which can achieve precise segmentation of gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) using a single model in T1-weighted imaging of infants aged 0-24 months.
Ultrasound Image-to-Video Synthesis via Latent Dynamic Diffusion Models
Chen, Tingxiu (MedAI Technology (Wuxi) Co Ltd), Mou, Lichao (Xidian University)
ClassificationGenerationData SynthesisDiffusion modelAuto EncoderImageVideoBiomedical DataUltrasound
🎯 What it does: A latent dynamic diffusion model (LDDM) is proposed, which can synthesize playable ultrasound videos from a single ultrasound image and use these synthesized videos for training video classification models, thereby alleviating the problem of scarcity of clinical ultrasound video data.
Uncertainty-aware Diffusion-based Adversarial Attack for Realistic Colonoscopy Image Synthesis
Jeong, Minjae (Pohang University of Science and Technology), Kim, Won Hwa (Pohang University of Science and Technology)
SegmentationData SynthesisAdversarial AttackDiffusion modelImageStochastic Differential Equation
🎯 What it does: This paper proposes a pixel-level uncertainty-aware adversarial attack data augmentation method based on diffusion models to enhance the robustness of polyp segmentation in colonoscopy.
Uncertainty-aware meta-weighted optimization framework for domain-generalized medical image segmentation
Oh, Seok-Hwan (KAIST), Bae, Hyeon-Min (KAIST)
SegmentationDomain AdaptationOptimizationMeta LearningDiffusion modelImageBiomedical DataUltrasound
🎯 What it does: A framework for cardiac image segmentation is proposed, which utilizes a diffusion generative model to synthesize diverse ultrasound images and optimizes through meta-learning spatial weighting, achieving domain generalization.
Uncertainty-Aware Multi-View Learning for Prostate Cancer Grading with DWI
Dong, Zhicheng (Shanghai University), Liang, Jiye (Tongji University)
ClassificationConvolutional Neural NetworkBiomedical DataMagnetic Resonance ImagingDiffusion Tensor Imaging
🎯 What it does: This paper proposes a multi-view learning method based on uncertainty perception, using diffusion-weighted imaging (DWI) with different b-values to achieve prostate cancer grading.
Uncovering Cortical Pathways of Prion-like Pathology Spreading in Alzheimer’s Disease by Neural Optimal Mass Transport
Huang, Yanquan (University of North Carolina at Chapel Hill), Wu, Guorong (POSTECH)
GenerationOptimizationExplainability and InterpretabilityGraph Neural NetworkGenerative Adversarial NetworkBiomedical DataPositron Emission TomographyAlzheimer's DiseaseOrdinary Differential Equation
🎯 What it does: An interpretable generative adversarial network based on neural optimal transport (OTM) is proposed to predict the propagation pathways of tau protein in the cortical region of the brain and its evolution over time in Alzheimer's disease.
Understanding Brain Dynamics Through Neural Koopman Operator with Structure-Function Coupling
Chow, Chiyuen (University of North Carolina at Chapel Hill), Wu, Guorong (POSTECH)
ClassificationGraph Neural NetworkTime SeriesBiomedical DataMagnetic Resonance Imaging
🎯 What it does: An end-to-end deep model named SKoop-C is proposed, which combines Geometric Scattering Transform (GST) with Neural Koopman Operator and a control module to simultaneously characterize the coupled dynamics of brain structure and function, and to perform cognitive task classification based on this.
Unified Multi-Modal Learning for Any Modality Combinations in Alzheimer’s Disease Diagnosis
Feng, Yidan (Hong Kong Polytechnic University), Qin, Jing (Hong Kong Polytechnic University)
TransformerMultimodalityBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease
🎯 What it does: A unified multimodal learning framework is proposed, capable of handling any combination of modalities, missing modalities, and new modalities, applied to Alzheimer's disease diagnosis.
Unified Prompt-Visual Interactive Segmentation of Clinical Target Volume in CT for Nasopharyngeal Carcinoma with Prior Anatomical Information
Khor, Hee Guan (Tsinghua University), Liao, Hongen (Tsinghua University)
SegmentationSupervised Fine-TuningPrompt EngineeringImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes SAM-RT, which combines multi-center and multi-modal anatomical prior knowledge with the SAM model for the segmentation of clinical target volumes (CTV) in nasopharyngeal carcinoma CT images.
uniGradICON: A Foundation Model for Medical Image Registration
Tian, Lin (University of North Carolina at Chapel Hill), Niethammer, Marc (UCSD)
SegmentationOptimizationConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: A general medical image registration foundational model called uniGradICON is proposed, which can achieve fast and accurate registration on images from various anatomical regions, modalities, and sources;
Universal Semi-Supervised Learning for Medical Image Classification
Ju, Lie (Monash University), Ge, Zongyuan (Melbourne University)
ClassificationDomain AdaptationConvolutional Neural NetworkAuto EncoderImageBiomedical Data
🎯 What it does: A unified framework is proposed to utilize unlabeled data from unknown categories and unknown domains for general semi-supervised learning in medical image classification.
Universal Topology Refinement for Medical Image Segmentation with Polynomial Feature Synthesis
Li, Liu (Imperial College London), Kainz, Bernhard (Imperial College London)
SegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A general topology refinement post-processing module is proposed, which enhances the topological accuracy of medical image segmentation through a plug-and-play network trained with topological perturbation masks synthesized by orthogonal polynomials.
Unsupervised Domain Adaptation using Soft-Labeled Contrastive Learning with Reversed Monte Carlo Method for Cardiac Image Segmentation
Gu, Mingxuan (Friedrich-Alexander-Universität Erlangen-Nürnberg), Maier, Andreas (Friedrich-Alexander-Universität Erlangen)
SegmentationDomain AdaptationContrastive LearningImageBiomedical DataMagnetic Resonance ImagingComputed Tomography
🎯 What it does: To address the problem of unsupervised domain adaptation for cardiac image segmentation, a contrastive learning framework based on soft pseudo-labels is proposed, combined with a reverse Monte Carlo method and centroid norm regularization to enhance feature clustering and alignment effects.
Unsupervised Latent Stain Adaptation for Computational Pathology
Reisenbüchler, Daniel (University of Regensburg), Merhof, Dorit (Hannover Medical School)
ClassificationSegmentationDomain AdaptationConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical Data
🎯 What it does: An unsupervised latent staining adaptation method ULSA is proposed, which jointly trains labeled target stained images synthesized by cGAN and unlabeled target stained images to achieve a staining-invariant model for computational pathology tasks.
Unsupervised Training of Neural Cellular Automata on Edge Devices
Kalkhof, John (Darmstadt University of Technology), Mukhopadhyay, Anirban (Carl Zeiss AG)
SegmentationImageBiomedical DataComputed Tomography
🎯 What it does: Implementing Neural Cellular Automata (NCA) for unsupervised segmentation and fine-tuning of lung X-ray images on Android phones.
Unsupervised Ultrasound Image Quality Assessment with Score Consistency and Relativity Co-learning
Guo, Juncheng (Hunan university), Li, Kenli (Hunan university)
ImageVideoUltrasound
🎯 What it does: An unsupervised fetal ultrasound image quality assessment method CRL-UIQA is proposed, which can output continuous perceptual quality scores for any standard plane.
UnWave-Net: Unrolled Wavelet Network for Compton Tomography Image Reconstruction
Ayad, Ishak (ETIS (UMR 8051), CY Cergy Paris University, ENSEA, CNRS), Nguyen, Maï K. (ETIS (UMR 8051), CY Cergy Paris University, ENSEA, CNRS)
RestorationOptimizationConvolutional Neural NetworkTransformerImageComputed Tomography
🎯 What it does: A deep unfolding network named UnWave-Net is proposed for non-projected circular Compton scattering tomography (NCCCST) image reconstruction.
URCDM: Ultra-Resolution Image Synthesis in Histopathology
Cechnicka, Sarah (Imperial College London), Kainz, Bernhard (Imperial College London)
GenerationData SynthesisSuper ResolutionDiffusion modelImageBiomedical DataStochastic Differential Equation
🎯 What it does: Utilizing the Unified Recursive Cascade Diffusion Model (URCDM) to generate high-resolution whole slide images (WSI), achieving coherent synthesis from low to high magnification across multiple scales, with the generated images maintaining realism at different magnification levels.
UrFound: Towards Universal Retinal Foundation Models via Knowledge-Guided Masked Modeling
Yu, Kai (Agency for Science, Technology and Research), Liu, Yong (Beijing University of Posts and Telecommunications)
Representation LearningTransformerSupervised Fine-TuningImageTextMultimodality
🎯 What it does: UrFound is proposed, a universal retinal foundation model capable of simultaneously processing CFP and OCT images, and learns cross-modal universal representations through knowledge-guided masked modeling.
Variational Field Constraint Learning for Degree of Coronary Artery Ischemia Assessment
Zhang, Qi (Shanghai Jiao Tong University), Gao, Zhifan (Chinese University of Hong Kong)
ImageBiomedical DataComputed Tomography
🎯 What it does: This paper proposes a Variational Field Constrained Learning Method (VFCLM) for assessing the fractional flow reserve (FFR) of coronary artery ischemia from digital subtraction angiography images.
VCLIPSeg: Voxel-wise CLIP-Enhanced model for Semi-Supervised Medical Image Segmentation
Li, Lei (Xiamen University), Li, Shaozi (Xiamen University)
SegmentationContrastive LearningImageBiomedical DataComputed Tomography
🎯 What it does: In the context of scarce annotations for medical image segmentation, VCLIPSeg is proposed to achieve semi-supervised segmentation through voxel-wise CLIP embeddings.
VDPF: Enhancing DVT Staging Performance Using a Global-Local Feature Fusion Network
Xie, Xiaotong (Affiliated Panyu Central Hospital Guangzhou Medical University), Huang, Yi (Affiliated Panyu Central Hospital Guangzhou Medical University)
ClassificationTransformerImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a deep vein thrombosis (DVT) staging system based on black blood magnetic resonance imaging (BTI), constructing the VDPF framework and achieving joint analysis of global images and local lesion images through a global-local feature fusion module.
Vertex Proportion Loss for Multi-Class Cell Detection from Label Proportions
Pacheco, Carolina (Johns Hopkins University), Haeffele, Benjamin (Johns Hopkins University)
ClassificationObject DetectionConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: This paper proposes a weakly supervised multi-class blood cell detection method that utilizes label proportion for learning and achieves cell localization and classification within the same framework.
VertFound: Synergizing Semantic and Spatial Understanding for Fine-grained Vertebrae Classification via Foundation Models
Tang, Jinzhou (Sun Yat-sen University), Zhao, Shen (Tianjin University of Technology)
ClassificationObject DetectionSegmentationTransformerContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Proposes the VertFound framework, which integrates the semantic understanding of CLIP and the spatial localization of SAM to achieve fine-grained vertebral classification in spinal images.
Vessel-aware aneurysm detection using multi-scale deformable 3D attention
Ceballos-Arroyo, Alberto M. (Northeastern University), Jiang, Huaizu (Brigham and Women's Hospital)
Object DetectionSegmentationConvolutional Neural NetworkTransformerImageBiomedical DataComputed Tomography
🎯 What it does: A method for detecting vascular-aware cerebral aneurysms based on multi-scale deformable 3D attention is proposed.
Vestibular schwannoma growth prediction from longitudinal MRI by time-conditioned neural fields
Chen, Yunjie (Leiden University Medical Center), Staring, Marius (Leiden University Medical Center)
Convolutional Neural NetworkRecurrent Neural NetworkSupervised Fine-TuningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Utilizing neural fields and a time-conditioned recurrent module to predict the morphological changes of vestibular schwannomas over time.
VideoCutMix: Temporal Segmentation of Surgical Videos in Scarce Data Scenarios
Dhanakshirur, Rohan Raju (Indian Institute of Technology Delhi), Arora, Chetan (PGIMER Chandigarh)
SegmentationOptical FlowVideo
🎯 What it does: This paper proposes a flow-consistent VideoCutMix video enhancement method to improve the performance of temporal action segmentation in data-scarce surgical videos.
Vision-Based Neurosurgical Guidance: Unsupervised Localization and Camera-Pose Prediction
Sarwin, Gary (ETH Zurich), Konukoglu, Ender (ETH Zurich)
Object DetectionPose EstimationTransformerAuto EncoderVideoBiomedical DataMagnetic Resonance Imaging
🎯 What it does: By using an unsupervised autoencoder, the anatomical structure detection boxes in endoscopic videos are embedded into a three-dimensional latent space to construct surgical paths and predict endoscopic angles for real-time neurosurgical navigation.
Visual-Textual Matching Attention for Lesion Segmentation in Chest Images
Bui, Phuoc-Nguyen, Choo, Hyunseung (Sungkyunkwan University)
SegmentationConvolutional Neural NetworkTransformerImageMultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a medical image segmentation network that integrates multi-scale features, convolution, and multi-head self-attention (MHSA, MHCA) modules, aiming to capture both fine-grained spatial information and global contextual relationships simultaneously.
VLSM-Adapter: Finetuning Vision-Language Segmentation Efficiently with Lightweight Blocks
Dhakal, Manish (Nepal Applied Mathematics and Informatics Institute), Khanal, Bishesh (Nepal Applied Mathematics and Informatics Institute)
SegmentationComputational EfficiencyTransformerSupervised Fine-TuningVision Language ModelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: The VLSM-Adapter module is proposed, which utilizes lightweight adapters to efficiently fine-tune medical image segmentation while keeping the pre-trained Vision-Language model frozen.
Volume-optimal persistence homological scaffolds of hemodynamic networks covary with MEG theta-alpha aperiodic dynamics
Nguyen, Nghi (Bar Ilan University), Shen, Li (University Of Pennsylvania)
MultimodalityBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Persistent homology is applied to the functional connectivity networks generated by fMRI in the Human Connectome Project to extract the volume-optimal persistent cycles and calculate their persistent centrality. Subsequently, these high-order topological features are aligned and compared with the aperiodic power changes of MEG signals in the theta-alpha (4-12 Hz) range, exploring their spatial distribution and variations under different attention tasks.
VolumeNeRF: CT Volume Reconstruction from a Single Projection View
Liu, Jiachen (Beihang University), Bai, Xiangzhi (Beihang University)
RestorationData SynthesisConvolutional Neural NetworkNeural Radiance FieldImageBiomedical DataComputed Tomography
🎯 What it does: A method for single-view X-ray reconstruction of CT volumes based on NeRF is proposed.
Volumetric Conditional Score-based Residual Diffusion Model for PET/MR Denoising
Yoon, Siyeop (Massachusetts General Hospital and Harvard Medical School), Li, Quanzheng (Massachusetts General Hospital and Harvard Medical School)
RestorationDiffusion modelScore-based ModelImageBiomedical DataMagnetic Resonance ImagingPositron Emission Tomography
🎯 What it does: A three-dimensional PET/MR denoising model based on Conditional Residual Diffusion (CSRD) is proposed, utilizing 3D patch training for efficient denoising.
Voxel Scene Graph for Intracranial Hemorrhage
Sanner, Antoine P. (Technical University of Darmstadt), Mukhopadhyay, Anirban (Carl Zeiss AG)
Object DetectionSegmentationConvolutional Neural NetworkRecurrent Neural NetworkImageBiomedical DataComputed Tomography
🎯 What it does: Detect and segment cerebral hemorrhage in non-contrast head CT scans and generate a 3D voxel scene map, with the model simultaneously learning the relationship between hemorrhage and brain structures to provide clinical decision support.
Weak-supervised Attention Fusion Network for Carotid Artery Vessel Wall Segmentation
Lei, Haijun (Shenzhen University), Lei, Baiying (Shenzhen University)
SegmentationConvolutional Neural NetworkTransformerSimultaneous Localization and MappingImageMagnetic Resonance Imaging
🎯 What it does: A weakly supervised attention fusion network based on CNN and Transformer was designed and implemented for precise segmentation of the carotid artery vessel wall.
Weakly Supervised Learning of Cortical Surface Reconstruction from Segmentations
Ma, Qiang (Imperial College London), Rueckert, Daniel (Technical University of Munich)
SegmentationGenerationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A CoSeg framework is proposed, utilizing weakly supervised brain tissue segmentation data to train a temporal attention network, achieving differential homeomorphic deformation of the initial cortical mesh for rapid and accurate cortical surface reconstruction.
Weakly Supervised Tooth Instance Segmentation on 3D Dental Models with Multi-Label Learning
Wang, Haoyu (Xian Jiaotong University), Ma, Jianhua (Pazhou Lab)
SegmentationGraph Neural NetworkPoint Cloud
🎯 What it does: A weakly supervised tooth instance segmentation network WS-TIS is proposed, which completes the segmentation of tooth instances in 3D dental models using multi-label learning with only 50% point-level annotations and thematic labels.
Weakly-supervised Medical Image Segmentation with Gaze Annotations
Zhong, Yuan (Chinese University of Hong Kong), Dou, Qi (Huazhong University of Science and Technology)
SegmentationConvolutional Neural NetworkImageBiomedical Data
🎯 What it does: This paper proposes the use of eye-tracking fixation data as weak supervision to achieve medical image segmentation through multi-layer networks and consistency regularization.
When 3D Partial Points Meets SAM: Tooth Point Cloud Segmentation with Sparse Labels
Liu, Yifan (Chinese University of Hong Kong), Yuan, Yixuan (Southern University of Science and Technology)
SegmentationRepresentation LearningTransformerPrompt EngineeringContrastive LearningPoint Cloud
🎯 What it does: In the task of tooth point cloud segmentation with extremely sparse labels (only one point labeled per tooth), the SAMTooth framework is proposed, which utilizes the prompt capability of SAM to automatically generate point prompts and projects the 2D masks generated by SAM back into 3D space, achieving high-precision segmentation through contrastive learning.
When Diffusion MRI Meets Diffusion Model: A Novel Deep Generative Model for Diffusion MRI Generation
Zhu, Xi (University of Electronic Science and Technology of China), Zhang, Fan (Harvard Medical School and Brigham and Women's Hospital)
GenerationData SynthesisDiffusion modelAuto EncoderImageBiomedical DataMagnetic Resonance ImagingDiffusion Tensor Imaging
🎯 What it does: This paper proposes a latent diffusion model (LDM) based on rotation-invariant spherical harmonics (RISH) features, achieving high-quality dMRI images equivalent to 7T from dMRI with a 3T magnetic field strength.
Whole Heart 3D+T Representation Learning Through Sparse 2D Cardiac MR Images
Zhang, Yundi (Technical University of Munich), Pan, Jiazhen (Technical University of Munich)
SegmentationRepresentation LearningTransformerAuto EncoderImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: Using a self-supervised mask reconstruction framework, we perform encoding learning on multi-view 2D+T CMR images and generate a unified 3D+T representation for cardiac phenotype prediction and whole heart segmentation.
WIA-LD2ND: Wavelet-based Image Alignment for Self-supervised Low-Dose CT Denoising
Zhao, Haoyu (Wuhan University), Yu, Rui (University of Louisville)
RestorationContrastive LearningImageComputed Tomography
🎯 What it does: A self-supervised low-dose CT denoising method WIA-LD2ND is proposed, which only uses positive dose CT images and includes a waveform image alignment module and frequency-aware multi-scale loss.
WiNet: Wavelet-based Incremental Learning for Efficient Medical Image Registration
Cheng, Xinxing (University of Birmingham), Duan, Jinming (University of Birmingham)
SegmentationOptimizationExplainability and InterpretabilityComputational EfficiencyConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an incremental learning network called WiNet based on wavelet transform for efficient and interpretable 3D medical image registration.
WsiCaption: Multiple Instance Generation of Pathology Reports for Gigapixel Whole-Slide Images
Chen, Pingyi (Zhejiang University), Yang, Lin (Westlake University)
GenerationTransformerLarge Language ModelImageText
🎯 What it does: Collected and constructed nearly 10,000 pairs of high-quality whole slide images and pathology reports (PathText), and proposed a multi-instance generation framework (MI-Gen) to achieve gigapixel-level automatic generation of pathology reports.
WSSADN: A Weakly Supervised Spherical Age-Disentanglement Network for Detecting Developmental Disorders with Structural MRI
Xue, Pengcheng (Nanjing University of Aeronautics and Astronautics), Wen, Xuyun (Nanjing University of Aeronautics and Astronautics)
ClassificationAnomaly DetectionConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: A weakly supervised spherical age decoupling network (WSSADN) is proposed, which extracts disease-related features for diagnosing developmental disorders by removing age information related to normal development in brain structures.
XA-Sim2Real: Adaptive Representation Learning for Vessel Segmentation in X-ray Angiography
Zhang, Baochang (Technical University of Munich), Navab, Nassir (Technische Universität München)
SegmentationDomain AdaptationRepresentation LearningConvolutional Neural NetworkGenerative Adversarial NetworkContrastive LearningImageComputed Tomography
🎯 What it does: A framework for unlabeled X-ray vascular segmentation, XA-Sim2Real, is proposed, achieving high-quality segmentation through the alignment of simulated images and adaptive representations.
XCoOp: Explainable Prompt Learning for Computer-Aided Diagnosis via Concept-guided Context Optimization
Bie, Yequan (Hong Kong University of Science and Technology), Chen, Hao (Harvard Medical School)
OptimizationExplainability and InterpretabilityTransformerPrompt EngineeringVision Language ModelContrastive LearningImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes an explainable prompt learning framework called XCoOp, which achieves interpretability and high performance in medical image diagnosis through multi-granularity soft and hard prompt alignment.
XTranPrune: eXplainability-aware Transformer Pruning for Bias Mitigation in Dermatological Disease Classification
Ghadiri, Ali (University of New South Wales), Song, Yang (University of New South Wales)
ClassificationExplainability and InterpretabilityTransformerImage
🎯 What it does: This paper proposes XTranPrune, an interpretable method for pruning visual Transformers (DeiT) to mitigate gender/skin color bias in skin disease classification.
Zero-shot Low-field MRI Enhancement via Denoising Diffusion Driven Neural Representation
Lin, Xiyue (ShanghaiTech University), Wei, Hongjiang (Shanghai Jiao Tong University)
RestorationDiffusion modelImageBiomedical DataMagnetic Resonance ImagingStochastic Differential Equation
🎯 What it does: A method based on a diffusion model-driven implicit neural representation (DiffDeuR) is designed for unsupervised recovery of low-field MRI to high-field equivalent images.
Zoom Pattern Signatures for Fetal Ultrasound Structures
Alsharid, Mohammad (University of Oxford), Noble, J. Alison (University of Oxford)
ClassificationRecognitionConvolutional Neural NetworkImageVideoUltrasound
🎯 What it does: The study utilizes the 'zoom mode' of the operator's magnification/reduction during ultrasound scanning to distinguish different fetal structures (CRL and NT), classifying based solely on the zoom sequence rather than the image content.