A Super-resolution Guided Network for Improving Automated Thyroid Nodule Segmentation. (December 2022)

• Record Type: Journal Article
• Authors: Lin, Xingtao; Zhou, Xiaogen; Tong, Tong; Nie, Xingqing; Wang, Luoyan; Zheng, Haonan; Li, Jing; Xue, Ensheng; Chen, Shun; Zheng, Meijuan; Chen, Cong; Jiang, Haiyan; Du, Min; Gao, Qinquan
• Is Part Of: Computer methods and programs in biomedicine. Volume 227(2022)
• Publication Date: 2022-12
• Subjects: Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2022.107186
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: We propose a novel two-stage framework based on super-resolution reconstruction to suppress noise and improve image quality for improving thyroid nodules segmentation, which outperforms exiting sate-of-the-art methods. Our proposed N-shape network for segmentation includes a multi-scale input layer, ASPP blocks, attention blocks and a PAC module. An ablation study has been performed to show the effectiveness and the benefits of each module in the segmentation task. We propose to utilize a multi-scale input layer and attention blocks. Different from other studies, we added an attention module to the position between every scale input layer and the last convolution block in each layer so that the network can pay more attention to the global information of every scale and remove the influence of speckle noise on the segmentation. We propose a novel parallel atrous convolution (PAC) module to encode the multi-scale semantic feature maps. The proposed PAC module is cascaded by four atrous convolution brunches to extract high-level semantic features from different receptive fields. Abstract: Background and Objective: A thyroid nodule is an abnormal lump that grows in the thyroid gland, which is the early symptom of thyroid cancer. In order to diagnose and treat thyroid cancer at the earliest stage, it is desired to characterize the nodule accurately. Ultrasound thyroid nodules segmentation is a challenging task due to the speckle noise, intensity heterogeneity, low contrast and low resolution. In this paper, we propose a novel framework to improve the accuracy of thyroid nodules segmentation. Methods: Different from previous work, a super-resolution reconstruction network is firstly constructed to upscale the resolution of the input ultrasound image. After that, our proposed N-shape network is utilized to perform the segmentation task. The guidance of super-resolution reconstruction network can make the high-frequency information of the input thyroid ultrasound image richer and more comprehensive than the original image. Our N-shape network consists of several atrous spatial pyramid pooling blocks, a multi-scale input layer, a U-shape convolutional network with attention blocks and a proposed parallel atrous convolution(PAC) module. These modules are conducive to capture context information at multiple scales so that semantic features can be fully utilized for lesion segmentation. Especially, our proposed PAC module is beneficial to further improve the segmentation by extracting high-level semantic features from different receptive fields. We use the UTNI-2021 dataset for model training, validating and testing. Results: The experimental results show that our proposed method achieve a Dice value of 91.9%, a mIoU value of 87.0%, a Precision value of 88.0%, a Recall value 83.7% and a F1-score value of 84.3%, which outperforms most state-of-the-art methods. Conclusions: Our method achieves the best performance on the UTNI-2021 dataset and provides a new way of ultrasound image segmentation. We believe that our method can provide doctors with reliable auxiliary diagnosis information in clinical practice.