基于ACGAN和迁移学习的骨显像分类方法

Bone scintigraphic classification method based on ACGAN and transfer learning

由于骨显像存在样本数量有限、类别不平衡的问题,导致骨显像分类存在较大困难。为提升骨显像的分类准确率,本文提出了一种基于结合辅助分类器的生成对抗网络(ACGAN)数据生成和迁移学习的骨显像分类方法。首先,为解决骨显像类别不平衡的问题,设计了一种MU-ACGAN模型。该模型以U-Net为生成器框架,同时结合密集残差连接和通道-空间注意力机制结构来提升骨显像细节特征生成,判别器通过密集残差注意力卷积块提取骨显像特征进行判别;然后,结合传统数据增强方式进一步扩充数据量;最后,设计了一种多尺度卷积神经网络提取骨显像不同尺度的特征,提升分类效果。在模型训练过程中,采用两阶段迁移学习方式,优化模型的初始化参数、解决过拟合的问题。实验结果表明,本文提出方法分类准确率达到了85.71%,有效缓解了小样本骨显像数据集分类准确率不高的问题。

Owing to the limited availability of samples and unbalanced categories of bone images,it is diffi⁃cult to classify these images.To improve the classification accuracy of bone images,this study developed a bone-image classification method based on auxiliary classifier generative adversarial network(ACGAN)data generation and transfer learning.First,an multi-attention U-Net-based ACGAN(MU-ACGAN)model was designed to address the imbalance of bone-image categories.The model uses U-Net as the gen⁃erator framework and combines dense residual connection and channel-spatial attention mechanism to im⁃prove the generation of bone-image detail features.The discriminator extracts bone-image features by us⁃ing a dense residual attention convolution block for discrimination.Next,the amount of data was further expanded via combination with traditional data enhancement methods.Finally,a multi-scale convolutional neural network was designed to extract the features at different scales of bone imaging so as to improve the classification effect.In the model training process,a two-stage transfer learning method was adopted to op⁃timize the initialization parameters of the model and address the problem of overfitting.Experimental re⁃sults indicate that the classification accuracy of the proposed method reaches 85.71%,effectively alleviat⁃ing the problem of low classification accuracy on small sample bone-image datasets.