Development and validation of an endoscopic images-based deep learning model for detection with nasopharyngeal malignancies

Cancer Communications

Bingzhong Jing¹  Caisheng He¹  Qiuyan Chen¹  Mingyuan Chen¹  Fang Qiu²  Peiyu Huang²  Shuhui Lv²  Lin Wang²  Wenze Qiu²  Hu Liang²  Yanqun Xiang²  Xiang Guo²  Xiong Zou²  Guoying Liu²  Yijun Hua²  Haoyuan Mo²  Chong Zhao³  Bin Li⁴  Liangru Ke⁵  Xing Lv⁶  Rui Sun⁶  Ying Sun⁶  Yahui Yu⁶  Kajia Cao⁶  Haiqiang Mai⁶  Weixiong Xia⁶  Ling Guo⁶  Chaofeng Li⁶  Shanshan Guo⁶  Kuiyuan Liu⁶  Jingjing Miao⁶  Xinjun Huang⁶  Linquan Tang⁶  Yishan Wu⁶  Wangzhong Li⁶  Donghua Luo⁶  Chaonan Qian⁶ 

[1] Department of Information, Sun Yat-Sen University Cancer Center, Guangzhou, P. R. China;Department of Nasopharyngeal Carcinoma, Sun Yat-Sen University Cancer Center, Guangzhou, P. R. China;Department of Radiology, Sun Yat-Sen University Cancer Center, Guangzhou, P. R. China;Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Guangzhou, P. R. China;Precision Medicine Center, Sun Yat-Sen University Cancer Center, Guangzhou, P. R. China;State Key Laboratory of Oncology in South China, Collaborative Innovation Center of Cancer Medicine, Guangzhou, P. R. China

关键词: Nasopharyngeal malignancy;    Deep learning;    Differential diagnosis;    Automatic segmentation;   

DOI  :  10.1186/s40880-018-0325-9

学科分类：肿瘤学

来源: Springer

【 摘 要 】

Due to the occult anatomic location of the nasopharynx and frequent presence of adenoid hyperplasia, the positive rate for malignancy identification during biopsy is low, thus leading to delayed or missed diagnosis for nasopharyngeal malignancies upon initial attempt. Here, we aimed to develop an artificial intelligence tool to detect nasopharyngeal malignancies under endoscopic examination based on deep learning. An endoscopic images-based nasopharyngeal malignancy detection model (eNPM-DM) consisting of a fully convolutional network based on the inception architecture was developed and fine-tuned using separate training and validation sets for both classification and segmentation. Briefly, a total of 28,966 qualified images were collected. Among these images, 27,536 biopsy-proven images from 7951 individuals obtained from January 1st, 2008, to December 31st, 2016, were split into the training, validation and test sets at a ratio of 7:1:2 using simple randomization. Additionally, 1430 images obtained from January 1st, 2017, to March 31st, 2017, were used as a prospective test set to compare the performance of the established model against oncologist evaluation. The dice similarity coefficient (DSC) was used to evaluate the efficiency of eNPM-DM in automatic segmentation of malignant area from the background of nasopharyngeal endoscopic images, by comparing automatic segmentation with manual segmentation performed by the experts. All images were histopathologically confirmed, and included 5713 (19.7%) normal control, 19,107 (66.0%) nasopharyngeal carcinoma (NPC), 335 (1.2%) NPC and 3811 (13.2%) benign diseases. The eNPM-DM attained an overall accuracy of 88.7% (95% confidence interval (CI) 87.8%–89.5%) in detecting malignancies in the test set. In the prospective comparison phase, eNPM-DM outperformed the experts: the overall accuracy was 88.0% (95% CI 86.1%–89.6%) vs. 80.5% (95% CI 77.0%–84.0%). The eNPM-DM required less time (40 s vs. 110.0 ± 5.8 min) and exhibited encouraging performance in automatic segmentation of nasopharyngeal malignant area from the background, with an average DSC of 0.78 ± 0.24 and 0.75 ± 0.26 in the test and prospective test sets, respectively. The eNPM-DM outperformed oncologist evaluation in diagnostic classification of nasopharyngeal mass into benign versus malignant, and realized automatic segmentation of malignant area from the background of nasopharyngeal endoscopic images.

【 授权许可】

CC BY