DyABD is the first benchmark dataset for abdominal muscle segmentation in dynamic MRIs featuring exercise-induced anatomical changes and pre/post-surgery scans, where existing models achieve an average Dice score of 0.82.
hub
Nature methods18(2), 203–211 (2021)
17 Pith papers cite this work. Polarity classification is still indexing.
17
Pith papers citing it
hub tools
citation-role summary
background 1
method 1
citation-polarity summary
representative citing papers
AnyTwoReg is a set-based groupwise registration method that achieves zero-shot generalization across variable-length and variable-contrast cardiac MRI sequences by using permutation-invariant feature aggregation.
WING generates synthetic CT from MRI/CBCT by decomposing the target into lung, soft-tissue, and bone windows, fusing them via a differentiable soft-fusion operator, and refining with a Transformer, achieving state-of-the-art results on SynthRAD2025.
MFASSL adds mirror-paired views, a lightweight Mirror-Fusion Attention module, and reflection-consistency losses to improve SSL on bilateral data with ~2.7% extra parameters.
Echo4DIR reconstructs continuous 4D cardiac geometry from sparse 2D echocardiography videos using implicit representations, epipolar feature fusion, self-supervised domain adaptation, and radial SDF alignment to achieve up to 98.35% Dice overlap.
A teacher–student semi-supervised framework with alignment-preserving patch mixing, position-aware text augmentation, and positional contrastive learning improves medical referring segmentation at low label ratios.
Transfer-aware data allocation derived from observed power-law scaling laws for asymmetric knowledge transfer in 3D medical imaging outperforms standard proportional sampling by up to 58% and generalizes to new budgets.
DiGSeg repurposes diffusion U-Nets as generalist segmentation learners by conditioning on image-mask latents and multi-scale CLIP text features, achieving strong cross-domain performance.
CT-guided voxel-wise regularization for the displacement field improves whole-body cross-tracer PET registration over global regularization baselines on a 296-patient dataset.
CHIS steers pretrained diffusion models to generate histopathology images aligned with input structural masks via frequency-domain structural initialization and wavelet-based textural modulation without any training on annotated data.
WoundFormer modifies SegFormer with a spatially-preserving multi-scale aggregation head for multi-class wound tissue segmentation, reporting 81.9% Dice on the WoundTissueSeg dataset with gains over baselines.
SWoMo decouples symbolic rule-based motion modeling via scene graphs from visual realism via diffusion models, trained through inverse pairing of real cataract surgery videos reconstructed in the simulator for sim-to-real translation.
Presents an SSM-based hierarchical feature learning method for medical point clouds that reports superior performance on classification, completion, and segmentation using a new dataset MedPointS.
CA-GCL combines global contrastive learning with permutation-invariant text augmentation to deliver zero-shot 3D medical abnormality detection that is more robust to prompt changes than prior FVLP methods.
Targeted data augmentations let single-sequence 3D spine segmentation models generalize to seven unseen CT and MRI datasets with 155% average Dice gain and almost no in-domain loss.
CoRE aligns image tokens to a hierarchical concept library to simulate clinical reasoning for expert routing and demand-based growth in continual brain lesion segmentation, achieving SOTA on 12 tasks.
A fine-tuned 3D foundation segmentation model combined with cross pseudo supervision achieves robust liver segmentation across labeled and unlabeled multi-phase, multi-vendor MRI without spatial registration.
citing papers explorer
-
DyABD: The Abdominal Muscle Segmentation in Dynamic MRI Benchmark
DyABD is the first benchmark dataset for abdominal muscle segmentation in dynamic MRIs featuring exercise-induced anatomical changes and pre/post-surgery scans, where existing models achieve an average Dice score of 0.82.
-
Set-Based Groupwise Registration for Variable-Length, Variable-Contrast Cardiac MRI
AnyTwoReg is a set-based groupwise registration method that achieves zero-shot generalization across variable-length and variable-contrast cardiac MRI sequences by using permutation-invariant feature aggregation.
-
WING: A Window-Prior-Based Generative Network with Gated Inception for Cross-Modality CT Synthesis
WING generates synthetic CT from MRI/CBCT by decomposing the target into lung, soft-tissue, and bone windows, fusing them via a differentiable soft-fusion operator, and refining with a Transformer, achieving state-of-the-art results on SynthRAD2025.
-
Mirror-Fusion Attention for Reflection-Aware Self-Supervised Representation Learning
MFASSL adds mirror-paired views, a lightweight Mirror-Fusion Attention module, and reflection-consistency losses to improve SSL on bilateral data with ~2.7% extra parameters.
-
Echo4DIR: 4D Implicit Heart Reconstruction from 2D Echocardiography Videos
Echo4DIR reconstructs continuous 4D cardiac geometry from sparse 2D echocardiography videos using implicit representations, epipolar feature fusion, self-supervised domain adaptation, and radial SDF alignment to achieve up to 98.35% Dice overlap.
-
Semi-MedRef: Semi-Supervised Medical Referring Image Segmentation with Cross-Modal Alignment
A teacher–student semi-supervised framework with alignment-preserving patch mixing, position-aware text augmentation, and positional contrastive learning improves medical referring segmentation at low label ratios.
-
Knowledge Transfer Scaling Laws for 3D Medical Imaging
Transfer-aware data allocation derived from observed power-law scaling laws for asymmetric knowledge transfer in 3D medical imaging outperforms standard proportional sampling by up to 58% and generalizes to new budgets.
-
Diffusion Model as a Generalist Segmentation Learner
DiGSeg repurposes diffusion U-Nets as generalist segmentation learners by conditioning on image-mask latents and multi-scale CLIP text features, achieving strong cross-domain performance.
-
CT-Guided Spatially-varying Regularization for Voxel-Wise Deformable Whole-Body PET Registration
CT-guided voxel-wise regularization for the displacement field improves whole-body cross-tracer PET registration over global regularization baselines on a 296-patient dataset.
-
Controllable Histopathology Image Synthesis with Training-free Structural Initialization and Textural Modulation
CHIS steers pretrained diffusion models to generate histopathology images aligned with input structural masks via frequency-domain structural initialization and wavelet-based textural modulation without any training on annotated data.
-
WoundFormer: Multi-Scale Spatial Feature Fusion for Multi-Class Wound Tissue Segmentation
WoundFormer modifies SegFormer with a spatially-preserving multi-scale aggregation head for multi-class wound tissue segmentation, reporting 81.9% Dice on the WoundTissueSeg dataset with gains over baselines.
-
SWoMo: Neuro-Symbolic World Model for Cataract Surgery Simulation
SWoMo decouples symbolic rule-based motion modeling via scene graphs from visual realism via diffusion models, trained through inverse pairing of real cataract surgery videos reconstructed in the simulator for sim-to-real translation.
-
Hierarchical Feature Learning for Medical Point Clouds via State Space Model
Presents an SSM-based hierarchical feature learning method for medical point clouds that reports superior performance on classification, completion, and segmentation using a new dataset MedPointS.
-
CA-GCL: Cross-Anatomy Global-Local Contrastive Learning for Robust 3D Medical Image Understanding
CA-GCL combines global contrastive learning with permutation-invariant text augmentation to deliver zero-shot 3D medical abnormality detection that is more robust to prompt changes than prior FVLP methods.
-
One Sequence to Segment Them All: Efficient Data Augmentation for CT and MRI Cross-Domain 3D Spine Segmentation
Targeted data augmentations let single-sequence 3D spine segmentation models generalize to seven unseen CT and MRI datasets with 155% average Dice gain and almost no in-domain loss.
-
CoRE: Concept-Reasoning Expansion for Continual Brain Lesion Segmentation
CoRE aligns image tokens to a hierarchical concept library to simulate clinical reasoning for expert routing and demand-based growth in continual brain lesion segmentation, achieving SOTA on 12 tasks.
-
Label-Efficient Cross-Modality Generalization for Liver Segmentation in Multi-Phase MRI
A fine-tuned 3D foundation segmentation model combined with cross pseudo supervision achieves robust liver segmentation across labeled and unlabeled multi-phase, multi-vendor MRI without spatial registration.