Automated Pulmonary Nodule Classification in Computed Tomography Images Using a Deep Convolutional Neural Network Trained by Generative Adversarial Networks

Table 2

Classification results for various numbers of images used for pretraining.

The number of generated images	Classification accuracy [%]
	Benign	Malignant
0	51.9	84.9
20,000	51.9	93.9
40,000	63.0	93.9
60,000	66.7	93.9
80,000	55.6	84.8
100,000	63.0	84.8