Continual Alignment for SAM: Rethinking Foundation Models for Medical Image Segmentation in Continual Learning

• URL: http://arxiv.org/abs/2511.17201v1
• Date: Fri, 21 Nov 2025 12:27:49 GMT
• Title: Continual Alignment for SAM: Rethinking Foundation Models for Medical Image Segmentation in Continual Learning
• Authors: Jiayi Wang, Wei Dai, Haoyu Wang, Sihan Yang, Haixia Bi, Jian Sun,
• Abstract summary: In medical image segmentation, heterogeneous privacy policies across institutions often make joint training on pooled datasets infeasible.<n>While the Segment Anything Model (SAM) offers strong zero-shot priors, its large parameter count and computational overhead challenge practical deployment.<n>This paper demonstrates that the SAM paradigm is highly promising once its computational efficiency and performance can be balanced.
• Score: 42.22450978339022
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In medical image segmentation, heterogeneous privacy policies across institutions often make joint training on pooled datasets infeasible, motivating continual image segmentation-learning from data streams without catastrophic forgetting. While the Segment Anything Model (SAM) offers strong zero-shot priors and has been widely fine-tuned across downstream tasks, its large parameter count and computational overhead challenge practical deployment. This paper demonstrates that the SAM paradigm is highly promising once its computational efficiency and performance can be balanced. To this end, we introduce the Alignment Layer, a lightweight, plug-and-play module which aligns encoder-decoder feature distributions to efficiently adapt SAM to specific medical images, improving accuracy while reducing computation. Building on SAM and the Alignment Layer, we then propose Continual Alignment for SAM (CA-SAM), a continual learning strategy that automatically adapts the appropriate Alignment Layer to mitigate catastrophic forgetting, while leveraging SAM's zero-shot priors to preserve strong performance on unseen medical datasets. Experimented across nine medical segmentation datasets under continual-learning scenario, CA-SAM achieves state-of-the-art performance. Our code, models and datasets will be released on \mbox{https://github.com/azzzzyo/Continual-Alignment-for-SAM.}

Related papers

• Boundary-Aware Test-Time Adaptation for Zero-Shot Medical Image Segmentation [12.159529070716824]
  BA-TTA-SAM is a test-time adaptation framework that enhances the zero-shot segmentation performance of SAM via test-time adaptation.<n>Our framework consistently outperforms state-of-the-art models in medical image segmentation.
  arXiv  Detail & Related papers  (2025-12-04T07:08:21Z)
• SAMora: Enhancing SAM through Hierarchical Self-Supervised Pre-Training for Medical Images [19.136303136685207]
  We propose SAMora, a framework that captures hierarchical medical knowledge at the image, patch, and pixel levels.<n> SAMora is compatible with various SAM variants, including SAM2, SAMed, and H-SAM.<n>It achieves state-of-the-art performance in both few-shot and fully supervised settings while reducing fine-tuning epochs by 90%.
  arXiv  Detail & Related papers  (2025-11-09T03:54:51Z)
• VesSAM: Efficient Multi-Prompting for Segmenting Complex Vessel [68.24765319399286]
  We present VesSAM, a powerful and efficient framework tailored for 2D vessel segmentation.<n>VesSAM integrates (1) a convolutional adapter to enhance local texture features, (2) a multi-prompt encoder that fuses anatomical prompts, and (3) a lightweight mask decoder to reduce jagged artifacts.<n>VesSAM consistently outperforms state-of-the-art PEFT-based SAM variants by over 10% Dice and 13% IoU.
  arXiv  Detail & Related papers  (2025-11-02T15:47:05Z)
• Learnable Prompting SAM-induced Knowledge Distillation for Semi-supervised Medical Image Segmentation [47.789013598970925]
  We propose a learnable prompting SAM-induced Knowledge distillation framework (KnowSAM) for semi-supervised medical image segmentation.<n>Our model outperforms the state-of-the-art semi-supervised segmentation approaches.
  arXiv  Detail & Related papers  (2024-12-18T11:19:23Z)
• Adapting Segment Anything Model for Unseen Object Instance Segmentation [70.60171342436092]
  Unseen Object Instance (UOIS) is crucial for autonomous robots operating in unstructured environments. We propose UOIS-SAM, a data-efficient solution for the UOIS task. UOIS-SAM integrates two key components: (i) a Heatmap-based Prompt Generator (HPG) to generate class-agnostic point prompts with precise foreground prediction, and (ii) a Hierarchical Discrimination Network (HDNet) that adapts SAM's mask decoder.
  arXiv  Detail & Related papers  (2024-09-23T19:05:50Z)
• ASPS: Augmented Segment Anything Model for Polyp Segmentation [77.25557224490075]
  The Segment Anything Model (SAM) has introduced unprecedented potential for polyp segmentation. SAM's Transformer-based structure prioritizes global and low-frequency information. CFA integrates a trainable CNN encoder branch with a frozen ViT encoder, enabling the integration of domain-specific knowledge.
  arXiv  Detail & Related papers  (2024-06-30T14:55:32Z)
• WSI-SAM: Multi-resolution Segment Anything Model (SAM) for histopathology whole-slide images [8.179859593451285]
  We present WSI-SAM, enhancing Segment Anything Model (SAM) with precise object segmentation capabilities for histopathology images. To fully exploit pretrained knowledge while minimizing training overhead, we keep SAM frozen, introducing only minimal extra parameters. Our model outperforms SAM by 4.1 and 2.5 percent points on a ductal carcinoma in situ (DCIS) segmentation tasks and breast cancer metastasis segmentation task.
  arXiv  Detail & Related papers  (2024-03-14T10:30:43Z)
• BLO-SAM: Bi-level Optimization Based Overfitting-Preventing Finetuning of SAM [37.1263294647351]
  We introduce BLO-SAM, which finetunes the Segment Anything Model (SAM) based on bi-level optimization (BLO) BLO-SAM reduces the risk of overfitting by training the model's weight parameters and the prompt embedding on two separate subsets of the training dataset. Results demonstrate BLO-SAM's superior performance over various state-of-the-art image semantic segmentation methods.
  arXiv  Detail & Related papers  (2024-02-26T06:36:32Z)
• BA-SAM: Scalable Bias-Mode Attention Mask for Segment Anything Model [65.92173280096588]
  We address the challenge of image resolution variation for the Segment Anything Model (SAM) SAM, known for its zero-shot generalizability, exhibits a performance degradation when faced with datasets with varying image sizes. We present a bias-mode attention mask that allows each token to prioritize neighboring information.
  arXiv  Detail & Related papers  (2024-01-04T15:34:44Z)
• How to Efficiently Adapt Large Segmentation Model(SAM) to Medical Images [15.181219203629643]
  Segment Anything (SAM) exhibits impressive capabilities in zero-shot segmentation for natural images. However, when applied to medical images, SAM suffers from noticeable performance drop. In this work, we propose to freeze SAM encoder and finetune a lightweight task-specific prediction head.
  arXiv  Detail & Related papers  (2023-06-23T18:34:30Z)

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.