基于DCGAN的脑膜瘤与听神经瘤检测模型优化方法研究

陈静聪, 冉凤伟, 章浩伟, 刘颖

Optimization Methodology for Meningioma and Acoustic Neuroma Detection Model Based on DCGAN

CHEN Jingcong, RAN Fengwei, ZHANG Haowei, LIU Ying

图4 DCGAN损失函数改进前后结果对比图. (a) DCGAN损失函数为交叉熵脑膜瘤生成样本；(b) DCGAN损失函数为最小二乘脑膜瘤生成样本；(c) DCGAN损失函数为交叉熵损失函数听神经瘤生成样本；(d) DCGAN损失函数为最小二乘听神经瘤生成样本

Fig. 4 Comparison of results before and after improvement of DCGAN loss function. (a) The DCGAN loss function is the cross entropy loss function for generating samples from meningiomas; (b) The DCGAN loss function is the least squares loss function for generating samples from meningiomas; (c) The DCGAN loss function is the cross entropy loss function for generating samples from acoustic neuromas; (d) The DCGAN loss function is the least squares loss function for generating samples from acoustic neuromas