Scaling In-Context Segmentation with Hierarchical Supervision

• URL: http://arxiv.org/abs/2604.12752v2
• Date: Wed, 22 Apr 2026 09:19:51 GMT
• Title: Scaling In-Context Segmentation with Hierarchical Supervision
• Abstract summary: In-context learning enables medical image segmentation models to adapt to new anatomical structures.<n>We propose PatchICL, a hierarchical framework that combines selective image patching with multi-level supervision.<n>Compared to UniverSeg, a strong global-attention baseline, PatchICL achieves competitive in-domain CT segmentation accuracy while reducing compute by 44% at $512times512$ resolution.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In-context learning (ICL) enables medical image segmentation models to adapt to new anatomical structures from limited examples, reducing the clinical annotation burden. However, standard ICL methods typically rely on dense, global cross-attention, which scales poorly with image resolution. While recent approaches have introduced localized attention mechanisms, they often lack explicit supervision on the selection process, leading to redundant computation in non-informative regions. We propose PatchICL, a hierarchical framework that combines selective image patching with multi-level supervision. Our approach learns to actively identify and attend only to the most informative anatomical regions. Compared to UniverSeg, a strong global-attention baseline, PatchICL achieves competitive in-domain CT segmentation accuracy while reducing compute by 44\% at $512\times512$ resolution. On 35 out-of-domain datasets spanning diverse imaging modalities, PatchICL outperforms the baseline on 6 of 13 modality categories, with particular strength on modalities dominated by localized pathology such as OCT and dermoscopy. Training and evaluation code are available at https://github.com/tidiane-camaret/ic_segmentation

Related papers

• Bridging Vision Foundation Model Priors with CLIP for Spatial-aware Few-shot Anomaly Detection in Medical Images [48.646448077284276]
  We propose Spatial-FAD, a framework for effective few-shot medical anomaly detection.<n>It improves lesion localization by combining VFM spatial priors with CLIP semantics.<n>Our method achieves an average improvement of over 11.4% in Dice score and 1.8% in AUC.
  arXiv  Detail & Related papers  (2026-09-11T05:21:01Z)
• MedSAD-CLIP: Supervised CLIP with Token-Patch Cross-Attention for Medical Anomaly Detection and Segmentation [1.055294253870608]
  Medical anomaly detection (MAD) and segmentation play a critical role in assisting clinical diagnosis.<n>Recent CLIP-based studies are promising for anomaly detection in zero-/few-shot settings.<n>We study supervised adaptation of CLIP for MAD under a realistic clinical setting.
  arXiv  Detail & Related papers  (2026-03-18T03:37:44Z)
• XAI-CLIP: ROI-Guided Perturbation Framework for Explainable Medical Image Segmentation in Multimodal Vision-Language Models [4.5236257764997205]
  XAI-CLIP is an ROI-guided perturbation framework for medical image segmentation.<n>It integrates language-informed region localization with medical image segmentation and applying targeted, region-aware perturbations.<n>XAI-CLIP achieves up to a 60% reduction in runtime, a 44.6% improvement in dice score, and a 96.7% increase in Intersection-over-Union.
  arXiv  Detail & Related papers  (2026-02-01T00:27:06Z)
• GCA-ResUNet: Medical Image Segmentation Using Grouped Coordinate Attention [3.6679095759171645]
  GCA-ResUNet is an efficient medical image segmentation framework equipped with a lightweight and plug-and-play Grouped Coordinate Attention (GCA) module.<n>Extensive experiments on two widely used benchmarks, Synapse and ACDC, demonstrate that GCA-ResUNet achieves Dice scores of 86.11% and 92.64%, respectively.
  arXiv  Detail & Related papers  (2025-12-30T05:13:20Z)
• TGC-Net: A Structure-Aware and Semantically-Aligned Framework for Text-Guided Medical Image Segmentation [56.09179939570486]
  We propose TGC-Net, a CLIP-based framework focusing on parameter-efficient, task-specific adaptations.<n>TGC-Net achieves state-of-the-art performance with substantially fewer trainable parameters, including notable Dice gains on challenging benchmarks.
  arXiv  Detail & Related papers  (2025-12-24T12:06:26Z)
• MAMBO-NET: Multi-Causal Aware Modeling Backdoor-Intervention Optimization for Medical Image Segmentation Network [51.68708264694361]
  Confusion factors can affect medical images, such as complex anatomical variations and imaging modality limitations.<n>We propose a multi-causal aware modeling backdoor-intervention optimization network for medical image segmentation.<n>Our method significantly reduces the influence of confusion factors, leading to enhanced segmentation accuracy.
  arXiv  Detail & Related papers  (2025-05-28T01:40:10Z)
• Dual-scale Enhanced and Cross-generative Consistency Learning for Semi-supervised Medical Image Segmentation [49.57907601086494]
  Medical image segmentation plays a crucial role in computer-aided diagnosis. We propose a novel Dual-scale Enhanced and Cross-generative consistency learning framework for semi-supervised medical image (DEC-Seg)
  arXiv  Detail & Related papers  (2023-12-26T12:56:31Z)
• Class Attention to Regions of Lesion for Imbalanced Medical Image Recognition [59.28732531600606]
  We propose a framework named textbfClass textbfAttention to textbfREgions of the lesion (CARE) to handle data imbalance issues. The CARE framework needs bounding boxes to represent the lesion regions of rare diseases. Results show that the CARE variants with automated bounding box generation are comparable to the original CARE framework.
  arXiv  Detail & Related papers  (2023-07-19T15:19:02Z)
• Contrastive Semi-supervised Learning for Domain Adaptive Segmentation Across Similar Anatomical Structures [21.54339967787734]
  We propose Contrastive Semi-supervised learning for Cross Anatomy Domain Adaptation. It adapts a model to segment similar structures in a target domain. It requires only limited annotations in the target domain.
  arXiv  Detail & Related papers  (2022-08-18T02:54:04Z)
• Inconsistency-aware Uncertainty Estimation for Semi-supervised Medical Image Segmentation [92.9634065964963]
  We present a new semi-supervised segmentation model, namely, conservative-radical network (CoraNet) based on our uncertainty estimation and separate self-training strategy. Compared with the current state of the art, our CoraNet has demonstrated superior performance.
  arXiv  Detail & Related papers  (2021-10-17T08:49:33Z)
• Few-shot Medical Image Segmentation using a Global Correlation Network with Discriminative Embedding [60.89561661441736]
  We propose a novel method for few-shot medical image segmentation. We construct our few-shot image segmentor using a deep convolutional network trained episodically. We enhance discriminability of deep embedding to encourage clustering of the feature domains of the same class.
  arXiv  Detail & Related papers  (2020-12-10T04:01:07Z)
• Max-Fusion U-Net for Multi-Modal Pathology Segmentation with Attention and Dynamic Resampling [13.542898009730804]
  The performance of relevant algorithms is significantly affected by the proper fusion of the multi-modal information. We present the Max-Fusion U-Net that achieves improved pathology segmentation performance. We evaluate our methods using the Myocardial pathology segmentation (MyoPS) combining the multi-sequence CMR dataset.
  arXiv  Detail & Related papers  (2020-09-05T17:24:23Z)

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.