CMCNet: Aligning Ultrasound Image Embeddings with Textual TI-RADS Representations for Fine-Grained Thyroid Classification

• URL: http://arxiv.org/abs/2608.13939v1
• Date: Fri, 14 Aug 2026 04:19:33 GMT
• Title: CMCNet: Aligning Ultrasound Image Embeddings with Textual TI-RADS Representations for Fine-Grained Thyroid Classification
• Abstract summary: We introduce the STN dataset of 600 thyroid nodules with paired transverse and longitudinal ultrasound images, bounding box annotations, and complete labels for all five TI-RADS feature categories.<n>We demonstrate that text embeddings derived from standardized feature descriptions form a stable surrogate representation for TI-RADS risk levels.<n> Experimental results show that this embedding alignment strategy is more data-efficient and robust than direct learning.
• Score: 9.908614103137422
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Ultrasound is the primary imaging modality for assessing thyroid nodules, and the ACR TI-RADS framework standardizes diagnosis through five ultrasound feature categories that are aggregated into five risk levels (TR1-TR5). Although widely adopted in clinical practice, most deep learning approaches focus on binary malignancy classification, while multi-class prediction and explicit utilization of feature-level supervision remain underexplored, largely due to limited annotated data. In this study, we introduce the STN dataset of 600 thyroid nodules with paired transverse and longitudinal ultrasound images, bounding box annotations, and complete labels for all five TI-RADS feature categories. Following the clinical decision process, we investigate how structured feature information can guide representation learning during training while requiring only images at inference. We demonstrate that text embeddings derived from standardized feature descriptions form a stable surrogate representation for TI-RADS risk levels. Based on this observation, we propose CMCNet, which aligns image embeddings to fixed textual embeddings via a Center-Margin Contrastive Loss that simultaneously promotes intra-class compactness and inter-class separation. Experimental results show that this embedding alignment strategy is more data-efficient and robust than direct multitask learning, and consistently outperforms InfoNCE, center loss, a strong multitask baseline, and a VQA-style multimodal model, particularly in imbalanced settings. The dataset is freely available at doi: 10.5281/zenodo.19125693 and the source code is available at: https://www.healthinformaticslab.org/supp/.

Related papers

• Prototype-Enhanced Multi-View Learning for Thyroid Nodule Ultrasound Classification [7.268080443895895]
  Existing deep learning methods often exhibit limited robustness and generalisation when deployed across different ultrasound devices or clinical environments.<n>We propose PEMV-thyroid, a Prototype-Enhanced Multi-View learning framework that accounts for data heterogeneity.<n>Experiments on multiple thyroid ultrasound datasets demonstrate that PEMV-thyroid consistently outperforms state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-03-30T11:37:26Z)
• Representation-Level Adversarial Regularization for Clinically Aligned Multitask Thyroid Ultrasound Assessment [7.520393692451524]
  Thyroid ultrasound is the first-line exam for assessing thyroid nodules and determining whether biopsy is warranted.<n> radiologists produce two outputs: a nodule contour for measurement and a TI-RADS risk category based on sonographic criteria.<n>We develop a clinically guided multitask framework that jointly predicts the nodule mask and TI-RADS category within a single model.
  arXiv  Detail & Related papers  (2026-03-22T07:14:43Z)
• Layout-Guided Controllable Pathology Image Generation with In-Context Diffusion Transformers [57.54843029965778]
  Controllable pathology image synthesis requires reliable regulation of spatial layout, tissue morphology, and semantic detail.<n>We propose In-Context Diffusion Transformer (IC-DiT), a layout-aware generative model that incorporates spatial layouts, textual descriptions, and visual embeddings into a unified diffusion transformer.<n>IC-DiT achieves higher fidelity, stronger spatial controllability, and better diagnostic consistency than existing methods.
  arXiv  Detail & Related papers  (2026-03-11T06:14:11Z)
• A Semantically Enhanced Generative Foundation Model Improves Pathological Image Synthesis [82.01597026329158]
  We introduce a Correlation-Regulated Alignment Framework for Tissue Synthesis (CRAFTS) for pathology-specific text-to-image synthesis.<n>CRAFTS incorporates a novel alignment mechanism that suppresses semantic drift to ensure biological accuracy.<n>This model generates diverse pathological images spanning 30 cancer types, with quality rigorously validated by objective metrics and pathologist evaluations.
  arXiv  Detail & Related papers  (2025-12-15T10:22:43Z)
• Adapting HFMCA to Graph Data: Self-Supervised Learning for Generalizable fMRI Representations [57.054499278843856]
  Functional magnetic resonance imaging (fMRI) analysis faces significant challenges due to limited dataset sizes and domain variability between studies.<n>Traditional self-supervised learning methods inspired by computer vision often rely on positive and negative sample pairs.<n>We propose adapting a recently developed Hierarchical Functional Maximal Correlation Algorithm (HFMCA) to graph-structured fMRI data.
  arXiv  Detail & Related papers  (2025-10-05T12:35:01Z)
• Hierarchical Corpus-View-Category Refinement for Carotid Plaque Risk Grading in Ultrasound [29.02957425057645]
  We propose a novel Corpus-View-Category Refinement Framework (CVC-RF)<n>CVC-RF processes information from Corpus-, View-, and Category-levels, enhancing model performance.<n> Experimental results indicate that CVC-RF effectively models global features via multi-level refinement.
  arXiv  Detail & Related papers  (2025-06-29T06:20:15Z)
• HDC: Hierarchical Distillation for Multi-level Noisy Consistency in Semi-Supervised Fetal Ultrasound Segmentation [2.964206587462833]
  A novel semi-supervised segmentation framework, called HDC, is proposed incorporating adaptive consistency learning with a single-teacher architecture.<n>The framework introduces a hierarchical distillation mechanism with two objectives: Correlation Guidance Loss for aligning feature representations and Mutual Information Loss for stabilizing noisy student learning.
  arXiv  Detail & Related papers  (2025-04-14T04:52:24Z)
• Eye-gaze Guided Multi-modal Alignment for Medical Representation Learning [65.54680361074882]
  Eye-gaze Guided Multi-modal Alignment (EGMA) framework harnesses eye-gaze data for better alignment of medical visual and textual features. We conduct downstream tasks of image classification and image-text retrieval on four medical datasets.
  arXiv  Detail & Related papers  (2024-03-19T03:59:14Z)
• Best of Both Worlds: Multimodal Contrastive Learning with Tabular and Imaging Data [7.49320945341034]
  We propose the first self-supervised contrastive learning framework to train unimodal encoders. Our solution combines SimCLR and SCARF, two leading contrastive learning strategies. We show the generalizability of our approach to natural images using the DVM car advertisement dataset.
  arXiv  Detail & Related papers  (2023-03-24T15:44:42Z)
• Data-Efficient Vision Transformers for Multi-Label Disease Classification on Chest Radiographs [55.78588835407174]
  Vision Transformers (ViTs) have not been applied to this task despite their high classification performance on generic images. ViTs do not rely on convolutions but on patch-based self-attention and in contrast to CNNs, no prior knowledge of local connectivity is present. Our results show that while the performance between ViTs and CNNs is on par with a small benefit for ViTs, DeiTs outperform the former if a reasonably large data set is available for training.
  arXiv  Detail & Related papers  (2022-08-17T09:07:45Z)
• Less is More: Adaptive Curriculum Learning for Thyroid Nodule Diagnosis [50.231954872304314]
  We propose an Adaptive Curriculum Learning framework, which adaptively discovers and discards the samples with inconsistent labels. We also contribute TNCD: a Thyroid Nodule Classification dataset.
  arXiv  Detail & Related papers  (2022-07-02T11:50:02Z)
• Multi-label Thoracic Disease Image Classification with Cross-Attention Networks [65.37531731899837]
  We propose a novel scheme of Cross-Attention Networks (CAN) for automated thoracic disease classification from chest x-ray images. We also design a new loss function that beyond cross-entropy loss to help cross-attention process and is able to overcome the imbalance between classes and easy-dominated samples within each class.
  arXiv  Detail & Related papers  (2020-07-21T14:37:00Z)

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.