# A dual-branch deep learning framework with Mask-Guided Attention for thyroid nodule classification in ultrasound images

## TL;DR

This paper introduces a new deep learning framework that improves thyroid nodule classification in ultrasound images by focusing on relevant lesion areas.

## Contribution

The novel dual-branch framework with mask-guided attention enhances classification by emphasizing diagnostically relevant regions.

## Key findings

- The proposed framework achieved superior classification accuracy on ultrasound datasets from three medical centers.
- The mask-guided feature enhancement module improves model robustness by suppressing irrelevant information.

## Abstract

Thyroid nodules are common, and accurate classification into benign or malignant types is essential for effective clinical management. Although high-resolution ultrasound is the primary diagnostic tool, its accuracy is limited by operator dependency. Recent advances in deep learning have shown promise for automated and objective assessment, but many existing methods lack focus on lesion-specific regions, compromising model robustness. To overcome these limitations, we propose a novel dual-branch deep learning framework that combines lesion segmentation and classification. A key feature of this framework is a nodule mask-guided feature enhancement module, which leverages probability masks from the segmentation branch to guide the classification branch toward diagnostically relevant regions while suppressing irrelevant information. Evaluated on ultrasound datasets from three medical centers, our approach demonstrates superior classification accuracy compared to baseline methods, highlighting its potential as a reliable computer-aided diagnosis tool for thyroid nodules.