Beyond Forgetting in Continual Medical Image Segmentation: A Comprehensive Benchmark Study

• URL: http://arxiv.org/abs/2605.06160v1
• Date: Thu, 07 May 2026 12:48:56 GMT
• Title: Beyond Forgetting in Continual Medical Image Segmentation: A Comprehensive Benchmark Study
• Abstract summary: We present a benchmark study of continual medical image segmentation.<n>We first define three clinically motivated scenarios, to respectively capture the cross-center domain shift, the incremental anatomical structure segmentation, and the cross-organ segmentation.<n>We then introduce an evaluation framework that measures not only general performance and forgetting, but also plasticity, forward generalizability, parameter efficiency, and replay burden.
• Score: 12.877570856116428
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Continual learning (CL) is essential for deploying medical image segmentation models in clinical environments where imaging domains, anatomical targets, and diagnostic tasks evolve over time. However, continual segmentation still faces three main challenges. First, the scenarios for this task remain insufficiently standardized for real-world clinical settings. Second, existing research has been primarily focused on mitigating forgetting, overlooking the other essential properties such as plasticity. Third, a benchmark work with comprehensive evaluation on existing methods is stll desirable. To address these gaps, we present such benchmark study of continual medical image segmentation. We first define three clinically motivated scenarios, namely Domain-CL, Class-CL, and Organ-CL, to respectively capture the cross-center domain shift, the incremental anatomical structure segmentation, and the cross-organ segmentation. We then introduce an evaluation framework that measures not only general performance and forgetting, but also plasticity, forward generalizability, parameter efficiency, and replay burden. The results, from extensive experiments with representative CL methods, showed that it was still challenging to develop a model that could satisfy all the requirements simultaneously. Nevertheless, these studies also suggested that the replay-based methods achieve the best overall balance between stability and plasticity, the parameter-isolation methods should be effective at reducing forgetting, though at the cost of increased model size, and the forward generalizability remain a significantly understudied aspect of this research field. Finally, we discuss related learning paradigms and outline future directions for continual medical image segmentation.

Related papers

• Bridging Restoration and Diagnosis: A Comprehensive Benchmark for Retinal Fundus Enhancement [13.241000354696903]
  EyeBench-V2 is a benchmark designed to bridge the gap between enhancement model performance and clinical utility.<n>Our benchmark provides a rigorous, task-oriented analysis of existing generative models.
  arXiv  Detail & Related papers  (2026-04-04T17:24:23Z)
• Bridging the Domain Divide: Supervised vs. Zero-Shot Clinical Section Segmentation from MIMIC-III to Obstetrics [6.451459735222387]
  We advance clinical section segmentation through three key contributions.<n>We curate a new de-identified, section-labeled obstetrics notes dataset.<n>We evaluate transformer-based supervised models for section segmentation on a curated subset of MIMIC-III.
  arXiv  Detail & Related papers  (2026-02-19T16:25:07Z)
• A Continual Learning-driven Model for Accurate and Generalizable Segmentation of Clinically Comprehensive and Fine-grained Whole-body Anatomies in CT [67.34586036959793]
  There is no fully annotated CT dataset with all anatomies delineated for training.<n>We propose a novel continual learning-driven CT model that can segment complete anatomies.<n>Our single unified CT segmentation model, CL-Net, can highly accurately segment a clinically comprehensive set of 235 fine-grained whole-body anatomies.
  arXiv  Detail & Related papers  (2025-03-16T23:55:02Z)
• Dynamically evolving segment anything model with continuous learning for medical image segmentation [50.92344083895528]
  We introduce EvoSAM, a dynamically evolving medical image segmentation model.<n>EvoSAM continuously accumulates new knowledge from an ever-expanding array of scenarios and tasks.<n>Experiments conducted by surgical clinicians on blood vessel segmentation confirm that EvoSAM enhances segmentation efficiency based on user prompts.
  arXiv  Detail & Related papers  (2025-03-08T14:37:52Z)
• MedCLIP-SAMv2: Towards Universal Text-Driven Medical Image Segmentation [2.2585213273821716]
  We introduce MedCLIP-SAMv2, a novel framework that integrates the CLIP and SAM models to perform segmentation on clinical scans.<n>Our approach includes fine-tuning the BiomedCLIP model with a new Decoupled Hard Negative Noise Contrastive Estimation (DHN-NCE) loss.<n>We also investigate using zero-shot segmentation labels within a weakly supervised paradigm to enhance segmentation quality further.
  arXiv  Detail & Related papers  (2024-09-28T23:10:37Z)
• Retrieval-augmented Few-shot Medical Image Segmentation with Foundation Models [17.461510586128874]
  We propose a novel method that adapts DINOv2 and Segment Anything Model 2 for retrieval-augmented few-shot medical image segmentation.<n>Our approach uses DINOv2's feature as query to retrieve similar samples from limited annotated data, which are then encoded as memories and stored in memory bank.
  arXiv  Detail & Related papers  (2024-08-16T15:48:07Z)
• MedCLIP-SAM: Bridging Text and Image Towards Universal Medical Image Segmentation [2.2585213273821716]
  We propose a novel framework, called MedCLIP-SAM, that combines CLIP and SAM models to generate segmentation of clinical scans. By extensively testing three diverse segmentation tasks and medical image modalities, our proposed framework has demonstrated excellent accuracy.
  arXiv  Detail & Related papers  (2024-03-29T15:59:11Z)
• Rethinking Semi-Supervised Medical Image Segmentation: A Variance-Reduction Perspective [51.70661197256033]
  We propose ARCO, a semi-supervised contrastive learning framework with stratified group theory for medical image segmentation. We first propose building ARCO through the concept of variance-reduced estimation and show that certain variance-reduction techniques are particularly beneficial in pixel/voxel-level segmentation tasks. We experimentally validate our approaches on eight benchmarks, i.e., five 2D/3D medical and three semantic segmentation datasets, with different label settings.
  arXiv  Detail & Related papers  (2023-02-03T13:50:25Z)
• Mine yOur owN Anatomy: Revisiting Medical Image Segmentation with Extremely Limited Labels [54.58539616385138]
  We introduce a novel semi-supervised 2D medical image segmentation framework termed Mine yOur owN Anatomy (MONA) First, prior work argues that every pixel equally matters to the model training; we observe empirically that this alone is unlikely to define meaningful anatomical features. Second, we construct a set of objectives that encourage the model to be capable of decomposing medical images into a collection of anatomical features.
  arXiv  Detail & Related papers  (2022-09-27T15:50:31Z)
• Self-supervised Answer Retrieval on Clinical Notes [68.87777592015402]
  We introduce CAPR, a rule-based self-supervision objective for training Transformer language models for domain-specific passage matching. We apply our objective in four Transformer-based architectures: Contextual Document Vectors, Bi-, Poly- and Cross-encoders. We report that CAPR outperforms strong baselines in the retrieval of domain-specific passages and effectively generalizes across rule-based and human-labeled passages.
  arXiv  Detail & Related papers  (2021-08-02T10:42:52Z)
• Robust Medical Instrument Segmentation Challenge 2019 [56.148440125599905]
  Intraoperative tracking of laparoscopic instruments is often a prerequisite for computer and robotic-assisted interventions. Our challenge was based on a surgical data set comprising 10,040 annotated images acquired from a total of 30 surgical procedures. The results confirm the initial hypothesis, namely that algorithm performance degrades with an increasing domain gap.
  arXiv  Detail & Related papers  (2020-03-23T14:35:08Z)

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.