A Deep Learning Framework for Thyroid Nodule Segmentation and Malignancy Classification from Ultrasound Images

2025-11-14

AI Summary

To address substantial inter-observer variability and poor model interpretability in thyroid nodule segmentation and malignancy risk prediction from ultrasound images, this paper proposes a two-stage end-to-end deep learning framework. In the first stage, TransUNet performs precise nodule segmentation; in the second stage, ResNet-18 classifies malignancy based exclusively on the segmented clinically relevant region. The framework enhances interpretability by explicitly focusing on anatomical structures and improves discriminative capability by implicitly integrating high-level visual features. Evaluated on 349 real-world ultrasound images via five-fold cross-validation, the method achieves an F1 score of 0.852—significantly outperforming a handcrafted-feature + random forest baseline (0.829). To our knowledge, this is the first fully automated, clinically viable pipeline that jointly addresses nodule detection, segmentation, and malignancy prediction while ensuring high performance, robustness, and clinical trustworthiness.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Segmentation
• Reasoning under Uncertainty: Uncertainty Representations

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Ultrasound-based risk stratification of thyroid nodules is a critical clinical task, but it suffers from high inter-observer variability. While many deep learning (DL) models function as "black boxes," we propose a fully automated, two-stage framework for interpretable malignancy prediction. Our method achieves interpretability by forcing the model to focus only on clinically relevant regions. First, a TransUNet model automatically segments the thyroid nodule. The resulting mask is then used to create a region of interest around the nodule, and this localised image is fed directly into a ResNet-18 classifier. We evaluated our framework using 5-fold cross-validation on a clinical dataset of 349 images, where it achieved a high F1-score of 0.852 for predicting malignancy. To validate its performance, we compared it against a strong baseline using a Random Forest classifier with hand-crafted morphological features, which achieved an F1-score of 0.829. The superior performance of our DL framework suggests that the implicit visual features learned from the localised nodule are more predictive than explicit shape features alone. This is the first fully automated end-to-end pipeline for both detecting thyroid nodules on ultrasound images and predicting their malignancy.

Problem

• Automated thyroid nodule segmentation and malignancy classification from ultrasound images
• Reducing inter-observer variability in ultrasound-based thyroid nodule risk assessment
• Developing interpretable deep learning framework for clinical nodule analysis

Innovation

• TransUNet segments thyroid nodules automatically
• ResNet-18 classifies malignancy using localized regions
• Two-stage framework ensures interpretable deep learning predictions