Abstract
ObjectivesTo evaluate the performance of a novel convolutional neural network (CNN) for the classification of typical perifissural nodules (PFN).MethodsChest CT data from two centers in the UK and The Netherlands (1668 unique nodules, 1260 individuals) were collected. Pulmonary nodules were classified into subtypes, including “typical PFNs” on-site, and were reviewed by a central clinician. The dataset was divided into a training/cross-validation set of 1557 nodules (1103 individuals) and a test set of 196 nodules (158 individuals). For the test set, three radiologically trained readers classified the nodules into three nodule categories: typical PFN, atypical PFN, and non-PFN. The consensus of the three readers was used as reference to evaluate the performance of the PFN-CNN. Typical PFNs were considered as positive results, and atypical PFNs and non-PFNs were grouped as negative results. PFN-CNN performance was evaluated using the ROC curve, confusion matrix, and Cohen’s kappa.ResultsInternal validation yielded a mean AUC of 91.9% (95% CI 90.6–92.9) with 78.7% sensitivity and 90.4% specificity. For the test set, the reader consensus rated 45/196 (23%) of nodules as typical PFN. The classifier-reader agreement (k = 0.62–0.75) was similar to the inter-reader agreement (k = 0.64–0.79). Area under the ROC curve was 95.8% (95% CI 93.3–98.4), with a sensitivity of 95.6% (95% CI 84.9–99.5), and specificity of 88.1% (95% CI 81.8–92.8).ConclusionThe PFN-CNN showed excellent performance in classifying typical PFNs. Its agreement with radiologically trained readers is within the range of inter-reader agreement. Thus, the CNN-based system has potential in clinical and screening settings to rule out perifissural nodules and increase reader efficiency.Key Points• Agreement between the PFN-CNN and radiologically trained readers is within the range of inter-reader agreement.• The CNN model for the classification of typical PFNs achieved an AUC of 95.8% (95% CI 93.3–98.4) with 95.6% (95% CI 84.9–99.5) sensitivity and 88.1% (95% CI 81.8–92.8) specificity compared to the consensus of three readers.
Highlights
Since the publication of the National Lung Screening Trial (NLST) demonstrated a reduction in lung cancer–related mortality of 20% compared to chest X-ray [1], and this has been confirmed more recently by European studies [2, 3], resulting in a significant amount of interest on lung cancer screening using computed tomography
The convolutional neural networks (CNNs) model for the classification of typical Perifissural nodules (PFN) achieved an AUC of 95.8% with 95.6%
427/ 1,557 (27.4%) nodules were labeled as typical PFNs, and 1,045/1,557 (67.1%) nodules were labeled as non-PFNs
Summary
Since the publication of the National Lung Screening Trial (NLST) demonstrated a reduction in lung cancer–related mortality of 20% compared to chest X-ray [1], and this has been confirmed more recently by European studies [2, 3], resulting in a significant amount of interest on lung cancer screening using computed tomography. Perifissural nodules (PFN) are a sub-group of small-tointermediate-sized solid nodules, frequently identified by radiologists reporting chest CT scans. PFNs are defined as nodules that are attached to fissures, and are homogenous and solid with smooth margins and an oval/lentiform or triangular shape [6]. They account for approximately 20–30% of all solid pulmonary nodules found in the lung cancer screening setting as well as incidentally in the clinical setting [6,7,8]. Automated and consistent identification of PFNs with their exclusion from further consideration could reduce the workload of radiologists and prevent unnecessary follow-up scans being performed
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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 call
