Skip to main content
Search

# Medical Image Segmentation

25 related articles

Subscribe to updates
  1. arXiv

    Brain-metastasis segmentation adapts to new hospitals, modalities and lesion classes: FGCARL reaches 68.28% mean Dice on BraTS-METS 2023

    Clinical images arrive in batches by institution, modality and annotated structure, so a segmentation model must absorb a new imaging distribution and new lesion classes without forgetting what it already learned. FGCARL puts bidirectional feature-statistics alignment and Fisher-weighted parameter constraints into one incremental framework that needs no replay of historical patient data. Moving from UCSF-BMSR to BraTS-METS 2023, where domain and classes shift together, FGCARL with Swin UNETR reaches 68.28% mean Dice across the three tumor regions of the new domain, against 60.94% for the strongest comparator, CLAMTS.

  2. arXiv

    Compute only where the signal is: Conformal Adaptation certifies sparse inference, holding Dice close on brain-tumor segmentation

    Medical and astronomical images are orders of magnitude larger than natural images, yet tumors and galaxies occupy a tiny fraction of the data, and uniform compression such as quantization or pruning can erode precision exactly there. Conformal Adaptation first screens candidates with a low-resolution model, calibrates the selection threshold with conformal risk control so the discarded cells carry at most a user-set expected fraction of true signal mass, then refines only the selected regions. Across brain-tumor segmentation, head-and-neck segmentation, calcium spike-rate inference and galaxy flux super-resolution, the pipeline cuts computation while keeping accuracy close, and VISTA3D reaches an 82.10x FLOPs speedup in CPU-only inference.

  3. arXiv

    Alibaba DAMO's HoloCT fuses CT reading and 3D segmentation in one model, surpassing all four radiologists in sensitivity and overall accuracy in a multicentre reader study

    Alibaba DAMO Academy's HoloCT is a 3D CT foundation model that puts scan interpretation and voxel-level segmentation inside one vision–language model, so a textual finding can come with a checkable 3D mask. Previously, segmentation networks delineated without interpreting, and CT vision–language models wrote text without voxel-level localization. On 11,455 clinical CT scans from five hospitals, HoloCT-8B raised mean Dice over VoxTell from 0.107 to 0.488 on laryngeal lesions and from 0.543 to 0.715 on lung cancer, and in retrospective reader studies of thyroid tumours and colon lesions it exceeded all four participating radiologists in sensitivity and overall accuracy.

  4. arXiv

    LungMend fills the cavities severe-TB CT masks miss, using only a healthy-lung shape prior and reaching 0.986 Dice

    On severe tuberculosis CT, cavities and consolidations have soft-tissue density and are left out of lung masks by segmenters such as R231. LungMend treats mask correction as shape completion: a latent diffusion prior trained on 57 healthy lungs, conditioned on a synthetically corrupted mask and on bone anatomy that disease does not alter, with no pathological data or CT intensities. On 200 radiologist-corrected slices from 20 severe TB cases it reaches 0.986 Dice and 0.57 mm average surface distance, with cavity and consolidation recall of 0.993 and 0.981.

  5. Medical contrastive learning stops forcing invariance: ProbSSL models embeddings as distributions and reaches 84.89 mean Dice on BTCV

    Medical self-supervised contrastive learning assumes two augmented views of the same image should share one semantic representation, but rotation and intensity perturbations can erase subtle yet critical anatomical and pathological cues, so rigid alignment suppresses task-relevant information. ProbSSL models each image embedding as a Gaussian, with the mean carrying shared semantics and the variance recording augmentation-induced variability, and aligns distributions by KL divergence so dimensions with larger variability receive weaker alignment pressure. After pre-training on CT, transfer to BTCV, MM-WHS, CC-CCII and BraTS gives higher average scores than the compared general and medical self-supervised methods on all four downstream tasks, including 84.89 mean Dice on BTCV.

  6. When hair breaks a lesion border, repair the high frequencies before decoding: 89.69% Dice on ISIC 2018

    In dermoscopic images, hair occlusion, uneven illumination and weak local contrast fragment lesion boundaries, while self-attention tends to emphasize lower-frequency content and leaves those weak borders underrepresented in skip connections. Swin-WaveNet decomposes each encoder skip feature with a wavelet transform, applies selective state-space scanning only to the high-frequency subbands to reconnect interrupted contours, gates shallow high-frequency responses with deeper semantic guidance to suppress hair and artifact structures, and sharpens boundaries during upsampling with high-frequency attention. On ISIC 2018 it reached a mean Dice of 89.69% and HD95 of 23.60 pixels over three runs, close to Swin-Unet, and 91.69% Dice on ISIC 2017 without fine-tuning.

  7. arXiv

    An MRI foundation model aligns by anatomy: no prompt box, zero-shot 3D organ localization matches MedSAM2

    MRI foundation models have largely learned image-level representations that struggle to separate normal from abnormal findings within one organ. A4-MR is pretrained in two stages on roughly 50k MRI studies and 491k series, pushing image-text alignment down to the anatomy level and learning separate embeddings for healthy and unhealthy anatomies. With a frozen encoder it gains 1.7 AUROC on abnormality detection over MedSigLIP, 1.4 Dice on knee segmentation over MedNeXt, stays within 0.4 Dice of the best model on abdominal organ segmentation, and localizes 3D anatomies without any prompt on par with MedSAM2 given a ground-truth box.

  8. Scoliosis CT labels without hand annotation: bending healthy spines into pathological shapes lifts vertebra Dice from 0.765 to 0.842

    Vertebra segmentation in severe scoliosis has been limited by scarce annotated pathological CT, since public data contain almost no scoliotic cases. An anatomy-aware synthesis pipeline jointly deforms public healthy spine CT scans and their segmentation labels into scoliosis-like volumes, then trains nnU-Net CNN and PRIMUS models under matched compute. On 18 in-house adolescent scoliosis CTs, the best CNN setting raised mean Dice from 0.765 to 0.842 and vertebra success rate from 73.2% to 81.6%; on 55 surgeon-defined apical labels, mean Dice rose from 0.720 to 0.854.

  9. arXiv

    MAGEFormer embeds voxel spacing into positional encoding and attention, reaching the lowest HD95 on anisotropic abdominal CT segmentation across BTCV and FLARE 22

    The authors propose MAGEFormer, which embeds physical spacing constraints into representation learning: MASE calibrates positional frequencies by voxel spacing, GCA suppresses physically implausible cross-scale feature correlations, and GVV performs deterministic sub-voxel view aggregation at inference; under a unified 5-fold protocol on BTCV and FLARE 22 it attains the strongest boundary accuracy among compared methods (10.58 mm and 3.40 mm HD95) with consistent Dice gains, and its advantage grows with anisotropy severity.

  10. Oxford team review: pericoronary fat imaging turns coronary inflammation into a quantifiable metric, with FAI Score falling after lipid-lowering, biologics and radiotherapy

    This review by a University of Oxford group synthesises the anatomical and physiological basis of pericoronary adipose tissue (PCAT) as a biosensor of coronary inflammation, explains how the standardised Fat Attenuation Index (FAI) Score corrects for technical, anatomical and biological variability and predicts MACE, reviews the role of artificial intelligence in automated segmentation and multi-parametric risk modelling, and compiles evidence that FAI falls after statins, anti-oxLDL antibodies, anti-TNF biologics, radiotherapy and cardiometabolic agents, concluding that PCAT imaging may complement traditional risk factors and plaque metrics while the evidence remains evolving.