1.
Application of different CNN models in AD diagnosis
不同 CNN 模型在 AD 诊断中的应用
模型 | 分类对象 | 影像数据 | 模型特点 | 准确率 | 相关文献 |
LeNet | AD/MCI | MRI/PET | 影像数据+ MMSE | 88.24% | [12] |
AlexNet | AD/MCI/NC | MRI | 修改卷积核参数 | 86.05% | [16] |
GoogleNet | AD/MCI/NC | PET | 数据增强+Adam | 83.23% | [19] |
VGGNet | AD/MCI/NC | MRI | VGG-16(DemNet) | 91.85% | [21] |
AD/MCI/NC | MRI | PFSECTL 模型 | 95.73% | [29] | |
ResNet | cMCI/sMCI | MRI | 大脑皮质数据 | 72.00% | [26] |
AD/MCI/NC | MRI | Inception 结构 | 81.00% | ||
RestNet-50 | 98.99% | [34] | |||
DenseNet | AD/MCI/NC | MRI | 概率集成模型 | 93.61% | [32] |