Anatomy-Guided Representation Learning Using a Transformer-Based Network for Thyroid Nodule Segmentation in Ultrasound Images

2025-12-14

AI Summary

Thyroid nodule ultrasound image segmentation faces three major challenges: ambiguous boundaries, highly variable nodule sizes, and severe scarcity of annotated data—leading to weak contextual modeling and poor generalization in existing models. To address these, we propose the first semi-supervised multi-task Transformer framework specifically designed for this task. Our method innovatively incorporates anatomical priors of the thyroid gland and jointly optimizes three complementary objectives: nodule segmentation, thyroid gland segmentation, and nodule size estimation. Key technical components include a hierarchical Transformer encoder, semi-supervised pretraining with consistency regularization, local-global feature fusion, and multi-task collaborative optimization—collectively enhancing boundary discrimination and scale robustness. On the TN3K and DDTI benchmarks, our approach achieves state-of-the-art Dice scores and superior cross-dataset generalization, demonstrating strong clinical applicability.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Computer Vision: Segmentation
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Accurate thyroid nodule segmentation in ultrasound images is critical for diagnosis and treatment planning. However, ambiguous boundaries between nodules and surrounding tissues, size variations, and the scarcity of annotated ultrasound data pose significant challenges for automated segmentation. Existing deep learning models struggle to incorporate contextual information from the thyroid gland and generalize effectively across diverse cases. To address these challenges, we propose SSMT-Net, a Semi-Supervised Multi-Task Transformer-based Network that leverages unlabeled data to enhance Transformer-centric encoder feature extraction capability in an initial unsupervised phase. In the supervised phase, the model jointly optimizes nodule segmentation, gland segmentation, and nodule size estimation, integrating both local and global contextual features. Extensive evaluations on the TN3K and DDTI datasets demonstrate that SSMT-Net outperforms state-of-the-art methods, with higher accuracy and robustness, indicating its potential for real-world clinical applications.

Problem

• Segment thyroid nodules in ultrasound images accurately
• Address ambiguous boundaries and size variations challenges
• Incorporate contextual information for better generalization

Innovation

• Transformer-based network for thyroid nodule segmentation
• Semi-supervised multi-task learning with unlabeled data
• Joint optimization of segmentation and size estimation tasks