Abstract
BackgroundTo develop and validate a deep learning–based model on CT images for the malignancy and invasiveness prediction of pulmonary subsolid nodules (SSNs).Materials and MethodsThis study retrospectively collected patients with pulmonary SSNs treated by surgery in our hospital from 2012 to 2018. Postoperative pathology was used as the diagnostic reference standard. Three-dimensional convolutional neural network (3D CNN) models were constructed using preoperative CT images to predict the malignancy and invasiveness of SSNs. Then, an observer reader study conducted by two thoracic radiologists was used to compare with the CNN model. The diagnostic power of the models was evaluated with receiver operating characteristic curve (ROC) analysis.ResultsA total of 2,614 patients were finally included and randomly divided for training (60.9%), validation (19.1%), and testing (20%). For the benign and malignant classification, the best 3D CNN model achieved a satisfactory AUC of 0.913 (95% CI: 0.885–0.940), sensitivity of 86.1%, and specificity of 83.8% at the optimal decision point, which outperformed all observer readers’ performance (AUC: 0.846±0.031). For pre-invasive and invasive classification of malignant SSNs, the 3D CNN also achieved satisfactory AUC of 0.908 (95% CI: 0.877–0.939), sensitivity of 87.4%, and specificity of 80.8%.ConclusionThe deep-learning model showed its potential to accurately identify the malignancy and invasiveness of SSNs and thus can help surgeons make treatment decisions.
Highlights
Lung cancer is one of the most lethal malignancies worldwide [1]
For the benign and malignant classification, the best 3D CNN model achieved a satisfactory area under the ROC curve (AUC) of 0.913, sensitivity of 86.1%, and specificity of 83.8% at the optimal decision point, which outperformed all observer readers’ performance (AUC: 0.846±0.031)
For pre-invasive and invasive classification of malignant subsolid nodules (SSNs), the 3D CNN achieved satisfactory AUC of 0.908, sensitivity of 87.4%, and specificity of 80.8%
Summary
Lung cancer is one of the most lethal malignancies worldwide [1]. Early detection and accurate diagnosis of pulmonary nodules can decrease the mortality of lung cancer [2]. According to the content of solid component, pulmonary nodules can be divided into solid nodules and subsolid nodules (SSNs) They have great difference in clinical management due to their different biological characteristics [3]. Malignant SSNs include subtypes of adenocarcinoma, and those malignant pathological types need careful intervention, such as surgical resection and stereotactic body radiation therapy (SBRT) [7]. The pre-invasive malignant SSNs may just need to be treated with conservative approach (sub-lobectomy or wedge resection) with long-term CT follow-up, while more aggressive surgical treatment (standard lobectomy with extended lymph node dissection) is necessary for patients with invasive (IA) SSNs. the prognosis of different pathological subtypes varies greatly after the corresponding treatment [10, 11]. To develop and validate a deep learning–based model on CT images for the malignancy and invasiveness prediction of pulmonary subsolid nodules (SSNs)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
