AUTHOR=Wang Luoyan , Zhou Xiaogen , Nie Xingqing , Lin Xingtao , Li Jing , Zheng Haonan , Xue Ensheng , Chen Shun , Chen Cong , Du Min , Tong Tong , Gao Qinquan , Zheng Meijuan 

TITLE=A Multi-Scale Densely Connected Convolutional Neural Network for Automated Thyroid Nodule Classification

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.878718

DOI=10.3389/fnins.2022.878718

ISSN=1662-453X

ABSTRACT=Automated thyroid nodule classification in ultrasound images is an important way to detect thyroid nodules and to make more accurate diagnosis. In this paper, we propose a novel deep convolutional neural network model, called n-ClsNet, for thyroid nodule classification. Our model consists of a multi-scale classification layer, multiple skip blocks and a hybrid atrous convolution block. The multi-scale classification layer first obtains multi-scale feature maps in order to make full use of image features. After that, each skip-block propagates information at different scales to learn multi-scale features for image classification. Finally, the hybrid atrous convolution block is used to replace the downpooling layer so that the spatial information can be fully learnt. We have evaluated our n-ClsNet model on the TNUI-2021 dataset. The proposed n-ClsNet achieves an average ACC score of 93.8% in the thyroid nodule classification task, which outperforms several representative state-of-the-art classification methods.