PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
October 16, 2025Machine Learning Science and Technology3 citationsOpen Access

Depthwise-dilated convolutional adapters for medical object tracking and segmentation using the segment anything model 2

View Full Paper
GXGuoping XuCKChristopher KabatYZYou Zhang

Key Points

  • DD-SAM2 helps fine-tune the Segment Anything Model 2 in medical videos, enhancing adaptability.
  • The use of depthwise-dilated adapters achieved Dice scores of 0.93 for tumor segmentation and 0.97 for left ventricle tracking.
  • The proposed method addresses the computational challenges of adapting deep learning models to medical imaging tasks.
  • Existing methods often fail with limited training data, but DD-SAM2 efficiently manages multi-scale feature extraction.

Abstract

Abstract Recent advances in medical image segmentation have been driven by deep learning; however, most existing methods remain limited by modality-specific designs and exhibit poor adaptability to dynamic medical imaging scenarios. The Segment Anything Model 2 (SAM2) and its related variants, which introduce a streaming memory mechanism for real-time video segmentation, present new opportunities for prompt-based, generalizable solutions. Nevertheless, adapting these models to medical video scenarios typically requires large-scale datasets for retraining or transfer learning, leading to high computational costs and the risk of catastrophic forgetting. To address these challenges, we propose DD-SAM2, an efficient adaptation framework for SAM2 that incorporates a Depthwise-Dilated Adapter (DD-Adapter) to enhance multi-scale feature extraction with minimal parameter overhead. This design enables effective fine-tuning of SAM2 on medical videos with limited training data. Unlike existing adapter-based methods focused solely on static images, DD-SAM2 fully exploits SAM2’s streaming memory for medical video objects tracking and segmentation. Comprehensive evaluations on TrackRad2025 (tumor segmentation) and EchoNet-Dynamic (left ventricle tracking) datasets demonstrate superior performance, achieving Dice scores of 0.93±0.04 and 0.97±0.01, respectively. To the best of our knowledge, this work provides an initial attempt at systematically exploring adapter-based SAM2 fine-tuning for medical video segmentation and tracking. Code, datasets, and models will be publicly available https://github.com/apple1986/DD-SAM2.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Xu et al. (2025) studied this question.

synapsesocial.com/papers/68f163c79903599108abcd20https://doi.org/10.1088/2632-2153/ae13d1
Ask AI
Helpful
Bookmark
Share
View Full Paper