Native-Space 3D CarveMix for Multi-Site T1w Stroke Segmentation

• URL: http://arxiv.org/abs/2608.23882v1
• Date: Mon, 24 Aug 2026 22:45:11 GMT
• Title: Native-Space 3D CarveMix for Multi-Site T1w Stroke Segmentation
• Abstract summary: We combine a MedNeXt-L backbone with on-the-fly 3D CarveMix augmentation that pastes real lesion patches into healthy brain regions during training.<n>Our method achieves a mean 5-fold cross-validation Dice of 0.648 at 500 epochs, a +0.018 improvement over the MedNeXt-L backbone at a matched training budget.
• Score: 0.11739570123067224
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Segmenting ischemic stroke lesions on T1-weighted (T1w) MRI acquired across different scanners and protocols without intensity standardization is difficult because lesions are subtle and share intensity characteristics with cerebrospinal fluid. Standard deep learning architectures trained across multiple centers plateau around Dice 0.66, with acute lesions ($\le 7$ days post-stroke) performing substantially worse due to severe sample scarcity. We combine a MedNeXt-L ($k=5$) backbone with on-the-fly 3D CarveMix augmentation that pastes real lesion patches into healthy brain regions during training. By generating synthetic lesion placements dynamically within each fold with subject-level split isolation, the model sees more diverse lesion patterns without requiring pre-generated copies on disk. We evaluate on 1,453 native T1w scans from 55 clinical centers in the ISLES 2026 challenge. Our method achieves a mean 5-fold cross-validation Dice of 0.648 at 500 epochs, a +0.018 improvement over the MedNeXt-L backbone at a matched training budget (0.630)

Related papers

• LeCor: Learning to Be Corrected by Meta-Learned Test-Time Training for Interactive 3D Lung-Tumour Segmentation [45.86925813888675]
  Fine-tuning SAM 3 on lung tumours raises the Dice obtained from a single point prompt from 0.298 to 0.757, and seven rounds of corrections raise it further to 0.765.<n>We propose LeCor, which performs test-time training on a small set of case adapters that are reset for every case.<n>LeCor raises the Dice reached after seven correction rounds from 0.787 with the fine-tuned model to 0.827, reduces the number of cases that never reach a Dice of 0.80 from 47 to 27, and reaches in three correction rounds the accuracy that the fine-tuned model attains in seven.
  arXiv  Detail & Related papers  (2026-09-08T21:49:56Z)
• PRISM-DR: Per-lesion Retinal Inference with Specialist Models for Diabetic Retinopathy [0.0]
  Diabetic retinopathy is a leading cause of preventable blindness.<n>Most automated detectors handle the four non-proliferative DR lesions.<n>We present PRISM-DR, a lesion-specific pipeline that trains one single-class detector per lesion.
  arXiv  Detail & Related papers  (2026-07-22T07:52:45Z)
• TriALS: Triphasic-Aided Liver Lesion Segmentation Benchmark in Non-Contrast CT [49.93685997940356]
  We describe the TriALS challenge for automated liver lesion segmentation under contrast-limited conditions.<n>The top-performing method achieved a mean venous-phase Dice of 0.754, consistent with human-level performance, yet dropped to 0.57 on NCCT.<n>Algorithm performance was most strongly predicted by training data scale and pre-training strategy.
  arXiv  Detail & Related papers  (2026-05-15T19:23:35Z)
• CrossPan: A Comprehensive Benchmark for Cross-Sequence Pancreas MRI Segmentation and Generalization [15.982441239969354]
  CrossPan is a benchmark comprising 1,386 3D scans across three routinely acquired sequences.<n>Models achieving Dice scores above 0.85 in-domain collapse to near-zero when transferred across sequences.<n>Semi-supervised learning offers gains only under stable intensity distributions and becomes unstable on generalization sequences with high intra-organ variability.
  arXiv  Detail & Related papers  (2026-04-20T20:00:32Z)
• C2W-Tune: Cavity-to -Wall Transfer Learning for Thin Atrial Wall Segmentation in 3D Late Gadolinium-enhanced Magnetic Resonance [1.452875650827562]
  C2W-Tune is a two-stage cavity-to-wall transfer framework.<n>It uses a high-accuracy LA cavity model as an anatomical prior to improve thin-wall delineation.<n>C2W-Tune achieved a Dice score of 0.78 and an HD95 of 3.15 mm, exceeding multi-class benchmarks.
  arXiv  Detail & Related papers  (2026-03-26T03:35:35Z)
• Hide-and-Seek Attribution: Weakly Supervised Segmentation of Vertebral Metastases in CT [68.09387763135236]
  We introduce a weakly supervised method trained solely on vertebra-level healthy/malignant labels, without any lesion masks.<n>We achieve strong blastic/lytic performance despite no mask supervision.
  arXiv  Detail & Related papers  (2025-12-07T14:03:28Z)
• MedNet-PVS: A MedNeXt-Based Deep Learning Model for Automated Segmentation of Perivascular Spaces [7.639673588124668]
  Enlarged perivascular spaces (PVS) are recognized as biomarkers of cerebral small vessel disease, Alzheimer's disease, stroke, and aging-related neurodegeneration.<n>We adapted MedNeXt-L-k5, a Transformer-inspired 3D encoder-decoder convolutional network, for automated PVS segmentation.
  arXiv  Detail & Related papers  (2025-08-27T20:24:12Z)
• MSWAL: 3D Multi-class Segmentation of Whole Abdominal Lesions Dataset [41.69818086021188]
  We introduce MSWAL, the first 3D Multi-class of the Whole Abdominal Lesions dataset.<n>MSWAL broadens the coverage of various common lesion types, such as gallstones, kidney stones, liver tumors, kidney tumors, pancreatic cancer, liver cysts, and kidney cysts.<n>We propose Inception nnU-Net, a novel segmentation framework that effectively integrates an Inception module with the nnU-Net architecture to extract information from different fields.
  arXiv  Detail & Related papers  (2025-03-17T06:31:25Z)
• Full-Head Segmentation of MRI with Abnormal Brain Anatomy: Model and Data Release [1.9492386975158917]
  We collected 98 MRIs with volumetric segmentation labels for a diverse set of human subjects.<n>We developed a MultiAxial network consisting of three 2D U-Net that operate independently in sagittal, axial and coronal planes.
  arXiv  Detail & Related papers  (2025-01-30T19:31:13Z)
• A Comprehensive Framework for Automated Segmentation of Perivascular Spaces in Brain MRI with the nnU-Net [37.179674347248266]
  Enlargement of perivascular spaces (PVS) is common in neurodegenerative disorders.<n>There is a need for reliable PVS detection methods which are currently lacking.
  arXiv  Detail & Related papers  (2024-11-29T09:19:57Z)
• Learning-based Bone Quality Classification Method for Spinal Metastasis [36.59899006688448]
  Early detection of spinal metastasis is critical for accurate staging and optimal treatment. In this paper, we explore a learning-based automatic bone quality classification method for spinal metastasis based on CT images.
  arXiv  Detail & Related papers  (2024-02-14T02:53:51Z)
• Segmentation of the Myocardium on Late-Gadolinium Enhanced MRI based on 2.5 D Residual Squeeze and Excitation Deep Learning Model [55.09533240649176]
  The aim of this work is to develop an accurate automatic segmentation method based on deep learning models for the myocardial borders on LGE-MRI. A total number of 320 exams (with a mean number of 6 slices per exam) were used for training and 28 exams used for testing. The performance analysis of the proposed ensemble model in the basal and middle slices was similar as compared to intra-observer study and slightly lower at apical slices.
  arXiv  Detail & Related papers  (2020-05-27T20:44:38Z)
• Handling Missing MRI Input Data in Deep Learning Segmentation of Brain Metastases: A Multi-Center Study [1.4463443378902883]
  A deep learning based segmentation model for automatic segmentation of brain metastases, named DropOut, was trained on multi-sequence MRI. The segmentation results were compared with the performance of a state-of-the-art DeepLabV3 model. The DropOut model showed a significantly higher score compared to the DeepLabV3 model.
  arXiv  Detail & Related papers  (2019-12-27T02:49:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.