Multi-Attribute Attention Network for Interpretable Diagnosis of Thyroid Nodules in Ultrasound Images

TL;DR

This paper introduces MAA-Net, a deep learning model that predicts thyroid nodule attributes and malignancy in ultrasound images, providing interpretable results aligned with clinical reasoning, and outperforms existing methods on a large dataset.

Contribution

The paper presents a novel multi-attribute attention network that mimics clinical diagnosis by learning attributes and using spatial priors, enhancing interpretability and accuracy.

Findings

Outperforms state-of-the-art methods on a large dataset

Provides interpretable, attribute-based malignancy predictions

Utilizes nodule delineations as spatial priors for better context understanding

Abstract

Ultrasound (US) is the primary imaging technique for the diagnosis of thyroid cancer. However, accurate identification of nodule malignancy is a challenging task that can elude less-experienced clinicians. Recently, many computer-aided diagnosis (CAD) systems have been proposed to assist this process. However, most of them do not provide the reasoning of their classification process, which may jeopardize their credibility in practical use. To overcome this, we propose a novel deep learning framework called multi-attribute attention network (MAA-Net) that is designed to mimic the clinical diagnosis process. The proposed model learns to predict nodular attributes and infer their malignancy based on these clinically-relevant features. A multi-attention scheme is adopted to generate customized attention to improve each task and malignancy diagnosis. Furthermore, MAA-Net utilizes nodule delineations as nodules spatial prior guidance for the training rather than cropping the nodules with additional models or human interventions to prevent losing the context information. Validation experiments were performed on a large and challenging dataset containing 4554 patients. Results show that the proposed method outperformed other state-of-the-art methods and provides interpretable predictions that may better suit clinical needs.