AI Navigate

Multi-Kernel Gated Decoder Adapters for Robust Multi-Task Thyroid Ultrasound under Cross-Center Shift

arXiv cs.CV / 3/11/2026

Ideas & Deep AnalysisModels & Research

Key Points

  • The paper addresses the challenge of automating thyroid ultrasound analysis, balancing global geometry-driven segmentation with local texture-driven malignancy assessment under cross-center domain shifts.
  • It identifies that convolutional neural networks (CNNs) and medical vision transformers (ViTs) exhibit complementary strengths, with ViTs better for geometric segmentation and CNNs preserving texture cues for malignancy under domain shift scenarios.
  • To mitigate negative transfer from shared backbones in multi-task learning, the authors propose Multi-Kernel Gated Adapters (MKGA) and a residual variant (ResMKGA) that refine features by selectively gating artifact-prone content using multi-scale and semantic context.
  • Experimental results on two ultrasound benchmarks demonstrate improved robustness in cross-center settings, enhancing out-of-domain segmentation and clinical diagnostic accuracy, notably with CNN backbones.
  • The authors plan to release their code and models to facilitate further research and practical adoption in medical imaging AI for thyroid ultrasound analysis.

Computer Science > Computer Vision and Pattern Recognition

arXiv:2603.08906 (cs)
[Submitted on 9 Mar 2026]

Title:Multi-Kernel Gated Decoder Adapters for Robust Multi-Task Thyroid Ultrasound under Cross-Center Shift

View a PDF of the paper titled Multi-Kernel Gated Decoder Adapters for Robust Multi-Task Thyroid Ultrasound under Cross-Center Shift, by Maziar Sabouri and 2 other authors
View PDF HTML (experimental)
Abstract:Thyroid ultrasound (US) automation couples two competing requirements: global, geometry-driven reasoning for nodule delineation and local, texture-driven reasoning for malignancy risk assessment. Under cross-center domain shift, these cues degrade asymmetrically, yet most multi-task pipelines rely on a single shared backbone, often inducing negative transfer. In this paper, we characterize this interference across CNN (ResNet34) and medical ViT (MedSAM) backbones, and observe a consistent trend: ViTs transfer geometric priors that benefit segmentation, whereas CNNs more reliably preserve texture cues for malignancy discrimination under strong shift and artifacts. Motivated by this failure mode, we propose a lightweight family of decoder-side adapters, the Multi-Kernel Gated Adapter (MKGA) and a residual variant (ResMKGA), which refine multi-scale skip features using complementary receptive fields and apply semantic, context-conditioned gating to suppress artifact-prone content before fusion. Across two US benchmarks, the proposed adapters improve cross-center robustness: they strengthen out-of-domain segmentation and, in the CNN setting, yield clear gains in clinical TI-RADS diagnostic accuracy compared to standard multi-task baselines. Code and models will be released.
Subjects: Computer Vision and Pattern Recognition (cs.CV); Medical Physics (physics.med-ph)
Cite as: arXiv:2603.08906 [cs.CV]
  (or arXiv:2603.08906v1 [cs.CV] for this version)
  https://doi.org/10.48550/arXiv.2603.08906
Focus to learn more
arXiv-issued DOI via DataCite

Submission history

From: Maziar Sabouri [view email]
[v1] Mon, 9 Mar 2026 20:18:35 UTC (7,797 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Multi-Kernel Gated Decoder Adapters for Robust Multi-Task Thyroid Ultrasound under Cross-Center Shift, by Maziar Sabouri and 2 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source
Current browse context:
cs.CV
< prev   |   next >
Change to browse by:

References & Citations

export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo
Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
Links to Code Toggle
Papers with Code (What is Papers with Code?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos

Demos

Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers

Recommenders and Search Tools

Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.