Lung Nodule Classification by the Combination of Fusion Classifier and Cascaded Convolutional Neural Networks

TL;DR

This paper introduces a combined approach using a fusion classifier and cascaded CNNs to improve lung nodule classification accuracy, especially addressing class imbalance issues, achieving high sensitivity at low false positive rates.

Contribution

It proposes a novel fusion classifier integrated with cascaded CNNs for better lung nodule classification in imbalanced datasets, enhancing detection sensitivity.

Findings

Achieved 94.4% sensitivity at 4 false positives per scan.

Achieved 95.9% sensitivity at 8 false positives per scan.

Demonstrated improved classification performance over previous methods.

Abstract

Lung nodule classification is a class imbalanced problem, as nodules are found with much lower frequency than non-nodules. In the class imbalanced problem, conventional classifiers tend to be overwhelmed by the majority class and ignore the minority class. We showed that cascaded convolutional neural networks can classify the nodule candidates precisely for a class imbalanced nodule candidate data set in our previous study. In this paper, we propose Fusion classifier in conjunction with the cascaded convolutional neural network models. To fuse the models, nodule probabilities are calculated by using the convolutional neural network models at first. Then, Fusion classifier is trained and tested by the nodule probabilities. The proposed method achieved the sensitivity of 94.4% and 95.9% at 4 and 8 false positives per scan in Free Receiver Operating Characteristics (FROC) curve analysis, respectively.