PURPOSE Deep convolutional neural networks (CNN) provide high accuracy for automatic classification of dopamine transporter (DAT) SPECT images. However, CNN are inherently black-box in nature lacking any kind of explanation for their decisions. This limits their acceptance for clinical use. This study tested layer-wise relevance propagation (LRP) to explain CNN-based classification of DAT-SPECT in patients with clinically uncertain parkinsonian syndromes. METHODS The study retrospectively included 1296 clinical DAT-SPECT with visual binary interpretation as "normal" or "reduced" by two experienced readers as standard-of-truth. A custom-made CNN was trained with 1008 randomly selected DAT-SPECT. The remaining 288 DAT-SPECT were used to assess classification performance of the CNN and to test LRP for explanation of the CNN-based classification. RESULTS Overall accuracy, sensitivity, and specificity of the CNN were 95.8%, 92.8%, and 98.7%, respectively. LRP provided relevance maps that were easy to interpret in each individual DAT-SPECT. In particular, the putamen in the hemisphere most affected by nigrostriatal degeneration was the most relevant brain region for CNN-based classification in all reduced DAT-SPECT. Some misclassified DAT-SPECT showed an "inconsistent" relevance map more typical for the true class label. CONCLUSION LRP is useful to provide explanation of CNN-based decisions in individual DAT-SPECT and, therefore, can be recommended to support CNN-based classification of DAT-SPECT in clinical routine. Total computation time of 3 s is compatible with busy clinical workflow. The utility of "inconsistent" relevance maps to identify misclassified cases requires further investigation.

中文翻译：

---

可解释的 AI 可提高卷积神经网络对多巴胺转运蛋白 SPECT 自动分类在临床不确定的帕金森综合征诊断中的接受度。

目的 深度卷积神经网络 (CNN) 为多巴胺转运蛋白 (DAT) SPECT 图像的自动分类提供高精度。然而，CNN 本质上是一个黑匣子，对他们的决定缺乏任何解释。这限制了它们对临床使用的接受。本研究测试了分层相关传播 (LRP)，以解释临床不确定的帕金森综合征患者中基于 CNN 的 DAT-SPECT 分类。方法 该研究回顾性地纳入了 1296 例临床 DAT-SPECT，由两名有经验的读者将视觉二元解释为“正常”或“减少”作为真理标准。使用 1008 个随机选择的 DAT-SPECT 训练定制的 CNN。其余 288 个 DAT-SPECT 用于评估 CNN 的分类性能并测试 LRP 以解释基于 CNN 的分类。结果 CNN 的总体准确度、敏感性和特异性分别为 95.8%、92.8% 和 98.7%。LRP 提供了易于在每个单独的 DAT-SPECT 中解释的相关图。特别是，受黑质纹状体变性影响最大的半球壳核是所有减少的 DAT-SPECT 中基于 CNN 分类的最相关的大脑区域。一些错误分类的 DAT-SPECT 显示了一个“不一致”的相关图，对于真实的类标签更为典型。结论 LRP 有助于解释单个 DAT-SPECT 中基于 CNN 的决策，因此可以推荐用于支持临床常规中基于 CNN 的 DAT-SPECT 分类。3 秒的总计算时间与繁忙的临床工作流程兼容。使用“不一致”相关图识别错误分类案例的效用需要进一步调查。