Convolutional Neural Network ensembles for accurate lung nodule malignancy prediction 2 years in the future. (July 2020)

• Record Type: Journal Article
• Authors: Paul, Rahul; Schabath, Matthew; Gillies, Robert; Hall, Lawrence; Goldgof, Dmitry
• Is Part Of: Computers in biology and medicine. Volume 122(2020)
• Publication Date: 2020-07
• Subjects: Radiomics -- Ensemble classification -- Convolutional Neural Network -- Lung Nodule -- NSCLC; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2020.103882
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: Convolutional Neural Networks (CNNs) have been utilized for to distinguish between benign lung nodules and those that will become malignant. The objective of this study was to use an ensemble of CNNs to predict which baseline nodules would be diagnosed as lung cancer in a second follow up screening after more than one year. Low-dose helical computed tomography images and data were utilized from the National Lung Screening Trial (NLST). The malignant nodules and nodule positive controls were divided into training and test cohorts. T0 nodules were used to predict lung cancer incidence at T1 or T2. To increase the sample size, image augmentation was performed using rotations, flipping, and elastic deformation. Three CNN architectures were designed for malignancy prediction, and each architecture was trained using seven different seeds to create the initial weights. This enabled variability in the CNN models which were combined to generate a robust, more accurate ensemble model. Augmenting images using only rotation and flipping and training with images from T0 yielded the best accuracy to predict lung cancer incidence at T2 from a separate test cohort (Accuracy = 90.29%; AUC = 0.96) based on an ensemble 21 models. Images augmented by rotation and flipping enabled effective learning by increasing the relatively small sample size. Ensemble learning with deep neural networks is a compelling approach that accurately predicted lung cancer incidence at the second screening after the baseline screen mostly 2 years later. Highlights: Three CNNs were designed using random initialization for ensemble prediction Two different image augmentation approaches were analyzed and compared. A snapshot ensemble, which requires less training time to build, was also evaluated. It's the best-known performance in predicting cancer malignancy 1 year in advance.