Pulmonary nodule risk classification in adenocarcinoma from CT images using deep CNN with scale transfer module

• Published in: IET image processing 2020-06, Vol.14 (8), p.1481-1489
• Main Authors: Zheng, Jie, Yang, Dawei, Zhu, Yu, Gu, Wanghuan, Zheng, Bingbing, Bai, Chunxue, Zhao, Lin, Shi, Hongcheng, Hu, Jie, Lu, Shaohua, Shi, Weibin, Wang, Ningfang
• Format: Article
• Language: English
• Subjects: authors; biological organs; cancer; classification accuracy; clinical treatment decision; computed tomography database; computerised tomography; CT images; deep CNN; deep convolutional neural network; different resolution images; early diagnosis; early‐stage; feature extraction; image classification; imaging studies; learning (artificial intelligence); lung; lung adenocarcinoma; lung adenocarcinomas risk; lung cancer; medical image processing; minimally invasive adenocarcinoma; named STM‐Net; neural nets; patient diagnosis; physicians diagnosis pulmonary nodules risk classification; pulmonary nodule risk classification; pulmonary nodules classification; Research Article; risk stage prediction; scale transfer module; Zhongshan Hospital Fudan University
• ISSN: 1751-9659; 1751-9667; 1751-9667
• DOI: 10.1049/iet-ipr.2019.0248

Pulmonary nodules risk classification in adenocarcinoma is essential for early detection of lung cancer and clinical treatment decision. Improving the level of early diagnosis and the identification of small lung adenocarcinoma has been always an important topic for imaging studies. In this study, the authors propose a deep convolutional neural network (CNN) with scale-transfer module (STM) and incorporate multi-feature fusion operation, named STM-Net. This network can amplify small targets and adapt to different resolution images. The evaluation data were obtained from the computed tomography (CT) database provided by Zhongshan Hospital Fudan University (ZSDB). All data have a pathological label and their lung adenocarcinomas risk are classified into four categories: atypical adenomatous hyperplasia, adenocarcinoma in situ, minimally invasive adenocarcinoma, and invasive adenocarcinoma. The authors’ deep learning network STM-Net was trained and tested for the risk stage prediction. The accuracy and the average area under the receiver operating characteristic curve achieved by their method are 95.455% and 0.987 for the ZSDB dataset. The experimental results show that STM-Net largely boosts classification accuracy on the pulmonary nodules classification compared with state-of-the-art approaches. The proposed method will be an effective auxiliary to help physicians diagnosis pulmonary nodules risk classification in adenocarcinoma in early-stage.