Fully-Automated Segmentation of Nasopharyngeal Carcinoma on Dual-Sequence MRI Using Convolutional Neural Networks.

Bin Huang,Bingsheng Huang,Yufeng Ye,Guorong Zou,Yan He,Zongyou Cai,Ping Zeng,Hanwei Chen,Wei Deng

doi:10.3389/fonc.2020.00166

Abstract

In this study, we proposed an automated method based on convolutional neural network (CNN) for nasopharyngeal carcinoma (NPC) segmentation on dual-sequence magnetic resonance imaging (MRI). T1-weighted (T1W) and T2-weighted (T2W) MRI images were collected from 44 NPC patients. We developed a dense connectivity embedding U-net (DEU) and trained the network based on the two-dimensional dual-sequence MRI images in the training dataset and applied post-processing to remove the false positive results. In order to justify the effectiveness of dual-sequence MRI images, we performed an experiment with different inputs in eight randomly selected patients. We evaluated DEU's performance by using a 10-fold cross-validation strategy and compared the results with the previous studies. The Dice similarity coefficient (DSC) of the method using only T1W, only T2W and dual-sequence of 10-fold cross-validation as different inputs were 0.620 ± 0.0642, 0.642 ± 0.118 and 0.721 ± 0.036, respectively. The median DSC in 10-fold cross-validation experiment with DEU was 0.735. The average DSC of seven external subjects was 0.87. To summarize, we successfully proposed and verified a fully automatic NPC segmentation method based on DEU and dual-sequence MRI images with accurate and stable performance. If further verified, our proposed method would be of use in clinical practice of NPC.

Highlights

Nasopharyngeal carcinoma (NPC) is a cancer type arising from the nasopharynx epithelium with a unique pattern of geographical distribution, with high incidence in Southeast Asia and North Africa [1]
Inspired by the successful application of deep convolutional neural network (CNN) in nasopharyngeal carcinoma (NPC) segmentation, in this study we proposed a dense connectivity embedding U-net (DEU) based on U-net, dense connectivity and dual-sequence magnetic resonance imaging (MRI) for accurate and automatic segmentation of NPC
We proposed an automated NPC segmentation method based on dual-sequence MRI images and CNN

Summary

Introduction

Nasopharyngeal carcinoma (NPC) is a cancer type arising from the nasopharynx epithelium with a unique pattern of geographical distribution, with high incidence in Southeast Asia and North Africa [1]. NPC has an incidence rate of 0.2‰ in endemic regions. The accurate delineation of NPC greatly influences radiotherapy planning. NPC cannot be clearly identified from the adjacent soft tissue on computed tomography (CT) image [3]. Compared with CT, magnetic resonance imaging (MRI) has demonstrated superior soft tissue contrast, Segmentation of NPC on MRI has been used as a preferred modality to evaluate the regional, local and intracranial infiltration of NPC. NPC has complex anatomical structure and often shares the similar intensities with the nearby tissues. NPC are delineated manually by radiologists or oncologists, which is time-consuming and subjective. Automatic segmentation methods can be faster and relatively objective

Methods

Results

Conclusion