Department of Cardiovascular Medicine, Nagasaki University Graduate School of Biomedical Sciences, Nagasaki, Japan.
Department of Radioisotope Medicine, Atomic Bomb Disease Institute, Nagasaki University, 1-7-1 Sakamoto, Nagasaki, 〒8528102, Japan.
Int J Cardiovasc Imaging. 2021 Jul;37(7):2337-2343. doi: 10.1007/s10554-021-02209-z. Epub 2021 Mar 11.
This study examined whether using an artificial neural network (ANN) helps beginners in diagnostic cardiac imaging to achieve similar results to experts when interpreting stress myocardial perfusion imaging (MPI). One hundred and thirty-eight patients underwent stress MPI with Tc-labeled agents. An expert and a beginner interpreted stress/rest MPI with or without the ANN and the results were compared. The myocardium was divided into 5 regions (the apex; septum; anterior; lateral, and inferior regions), and the defect score of myocardial blood flow was evaluated from 0 to 4, and SSS, SRS, and SDS were calculated. The ANN effect, defined as the difference in each of these scores between with and without the ANN, was calculated to investigate the influence of ANN on the interpreters' performance. We classified 2 groups (insignificant perfusion group and significant perfusion group) and compared them. In the same way, classified 2 groups (insignificant ischemia group and significant ischemia group) and compared them. Besides, we classified 2 groups (normal vessels group and multi-vessels group) and compared them. The ANN effect was smaller for the expert than for the beginner. Besides, the ANN effect for insignificant perfusion group, insignificant ischemia group and multi-vessels group were smaller for the expert than for the beginner. On the other hand, the ANN effect for significant perfusion group, significant ischemia group and normal vessels group were no significant. When interpreting MPI, beginners may achieve similar results to experts by using an ANN. Thus, interpreting MPI with ANN may be useful for beginners. Furthermore, when beginners interpret insignificant perfusion group, insignificant ischemia group and multi-vessel group, beginners may achieve similar results to experts by using an ANN.
本研究旨在探讨在解读应激心肌灌注成像（MPI）时，使用人工神经网络（ANN）是否有助于初学者达到与专家相似的诊断结果。138 名患者接受了 Tc 标记药物的应激 MPI。一位专家和一位初学者在有或没有 ANN 的情况下解读应激/静息 MPI，并比较结果。心肌分为 5 个区域（心尖；室间隔；前壁；侧壁和下壁），并从 0 到 4 评估心肌血流缺损评分，计算 SSS、SRS 和 SDS。计算 ANN 效果（定义为有无 ANN 时每个分数的差异），以研究 ANN 对解释者表现的影响。我们将患者分为 2 组（灌注无明显异常组和灌注明显异常组）并进行比较。同样地，将患者分为 2 组（缺血无明显异常组和缺血明显异常组）并进行比较。此外，我们将患者分为 2 组（正常血管组和多血管组）并进行比较。与初学者相比，专家的 ANN 效果较小。此外，对于灌注无明显异常组、缺血无明显异常组和多血管组，专家的 ANN 效果较小。另一方面，对于灌注明显异常组、缺血明显异常组和正常血管组，ANN 效果无显著差异。在解读 MPI 时，初学者可以通过使用 ANN 获得与专家相似的结果。因此，使用 ANN 解读 MPI 可能对初学者有用。此外，当初学者解读灌注无明显异常组、缺血无明显异常组和多血管组时，他们可以通过使用 ANN 获得与专家相似的结果。
