Multi-stage Neural Networks with Single-sided Classifiers for False Positive Reduction and its Evaluation using Lung X-ray CT Images

TL;DR

This paper introduces a multi-stage neural network approach with single-sided classifiers to effectively reduce false positives in lung nodule detection from CT images, addressing class imbalance and improving detection sensitivity.

Contribution

The paper proposes a cascaded CNN framework that filters obvious non-nodules before applying a balanced CNN, enhancing false positive reduction in lung nodule classification.

Findings

Achieved 92.4% sensitivity at 4 false positives per scan.

Achieved 94.5% sensitivity at 8 false positives per scan.

Demonstrated effectiveness on lung X-ray CT images.

Abstract

Lung nodule classification is a class imbalanced problem because 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 therefore propose cascaded convolutional neural networks to cope with the class imbalanced problem. In the proposed approach, multi-stage convolutional neural networks that perform as single-sided classifiers filter out obvious non-nodules. Successively, a convolutional neural network trained with a balanced data set calculates nodule probabilities. The proposed method achieved the sensitivity of 92.4\% and 94.5% at 4 and 8 false positives per scan in Free Receiver Operating Characteristics (FROC) curve analysis, respectively.