文献速递心脏核医学

基于深度卷积自编码器的内容图像检索用于心肌灌注成像诊断

Content-based image retrieval for the diagnosis of myocardial perfusion imaging using a deep convolutional autoencoder.
PubMed 文献 · J Nucl Cardiol · 2022-07-08

该研究从本院2019年12月至2022年2月连续接受心脏闪烁扫描的患者中收集843对负荷与静息心肌灌注图像,训练深度卷积自编码器(CAE)提取256维特征向量,经主成分分析(PCA)可视化后基于余弦相似度进行内容图像检索(CBIR)。PCA三维散点图显示特征向量保留了放射科医师报告的% summed difference score、缺血存在及瘢痕位置等临床信息;CBIR作为基于相似度的诊断工具时二元准确率为81.0%。结论提示无监督特征学习在MPI的CBIR中具有实用性。

要点

  • CAE提取的256维特征向量经PCA可视化后保留了放射科医师报告的缺血及瘢痕位置等临床信息。
  • 基于余弦相似度的CBIR作为相似度诊断工具时二元准确率为81.0%。
  • 结论提示无监督特征学习用于MPI的内容图像检索具有实用性。

核心结论

  • 结果表明无监督特征学习用于心肌灌注成像的内容图像检索具有实用性。

研究设计与样本

  • 从本院2019年12月至2022年2月连续接受心脏闪烁扫描的患者中收集843对负荷与静息心肌灌注图像。
  • 训练CAE模型以复现输入的配对图像数据,使编码器输出256维特征向量。
  • 提取的特征向量经PCA进一步降维用于数据可视化。
  • 基于查询图像与参考图像特征向量间的余弦相似度进行CBIR。
  • 采用二元准确率、精确率、召回率和F1分数评估放射科医师所见在查询与检索MPI之间的一致性。

主要发现

  • PCA三维散点图显示特征向量保留了放射科医师报告的% summed difference score、缺血存在及瘢痕位置等临床信息。
  • CBIR作为基于相似度的诊断工具时,二元准确率为81.0%。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) plays a crucial role in the optimal treatment strategy for patients with coronary heart disease. We tested the feasibility of feature extraction from MPI using a deep convolutional autoencoder (CAE) model. METHODS: Eight hundred and forty-three pairs of stress and rest myocardial perfusion images were collected from consecutive patients who underwent cardiac scintigraphy in our hospital between December 2019 and February 2022. We trained a CAE model to reproduce the input paired image data, so as the encoder to output a 256-dimensional feature vector. The extracted feature vectors were further dimensionally reduced via principal component analysis (PCA) for data visualization. Content-based image retrieval (CBIR) was performed based on the cosine similarity of the feature vectors between the query and reference images. The agreement of the radiologist's finding between the query and retrieved MPI was evaluated using binary accuracy, precision, recall, and F1-score. RESULTS: A three-dimensional scatter plot with PCA revealed that feature vectors retained clinical information such as percent summed difference score, presence of ischemia, and the location of scar reported by radiologists. When CBIR was used as a similarity-based diagnostic tool, the binary accuracy was 81.0%. CONCLUSION: The results indicated the utility of unsupervised feature learning for CBIR in MPI.

来源:J Nucl Cardiol,查看原文。版权归原作者所有。

引用

  1. Higaki A, Kawaguchi N, Kurokawa T, et al. Content-based image retrieval for the diagnosis of myocardial perfusion imaging using a deep convolutional autoencoder. J Nucl Cardiol. 2022. PMID: 35802346. doi:10.1007/s12350-022-03030-4

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。