Urs Meghana, Killekar Aditya, Builoff Valerie, Lemley Mark, Wei Chih-Chun, Ramirez Giselle, Kavanagh Paul, Buckley Christopher, Slomka Piotr J
Department of Medicine (Division of Artificial Intelligence in Medicine), Cedars-Sinai Medical Center, Los Angeles, CA.
GE Healthcare, Pharmaceutical Diagnostics, Buckinghamshire, UK.
medRxiv. 2025 Jul 1:2025.06.27.25330436. doi: 10.1101/2025.06.27.25330436.
Precise quantification of myocardial blood flow (MBF) and flow reserve (MFR) in F-flurpiridaz PET significantly relies on motion correction (MC). However, the manual frame-by-frame correction leads to significant inter-observer variability, time-consuming, and requires significant experience. We propose a deep learning (DL) framework for automatic MC of F-flurpiridaz PET.
The method employs a 3D ResNet based architecture that takes 3D PET volumes and outputs motion vectors. It was validated using 5-fold cross-validation on data from 32 sites of a Phase III clinical trial (NCT01347710). Manual corrections from two experienced operators served as ground truth, and data augmentation using simulated vectors enhanced training robustness. The study compared the DL approach to both manual and standard non-AI automatic MC methods, assessing agreement and diagnostic accuracy using minimal segmental MBF and MFR.
The area under the receiver operating characteristic curves (AUC) for significant CAD were comparable between DL-MC MBF, manual-MC MBF from Operators (AUC=0.897, 0.892 and 0.889, respectively; p>0.05), standard non-AI automatic MC (AUC=0.877; p>0.05) and significantly higher than No-MC (AUC=0.835; p<0.05). Similar findings were observed with MFR. The 95% confidence limits for agreement with the operator were ±0.49ml/g/min (mean difference = 0.00) for MFR and ±0.24ml/g/min (mean difference = 0.00) for MBF.
DL-MC is significantly faster but diagnostically comparable to manual-MC. The quantitative results obtained with DL-MC for MBF and MFR are in excellent agreement with those manually corrected by experienced operators compared to standard non-AI automatic MC in patients undergoing F-flurpiridaz PET-MPI.
在F-氟吡哒唑正电子发射断层扫描（PET）中，心肌血流量（MBF）和血流储备（MFR）的精确量化在很大程度上依赖于运动校正（MC）。然而，逐帧手动校正会导致显著的观察者间变异性，耗时且需要丰富经验。我们提出了一种用于F-氟吡哒唑PET自动MC的深度学习（DL）框架。
该方法采用基于3D ResNet的架构，以3D PET体积为输入并输出运动向量。使用来自一项III期临床试验（NCT01347710）32个站点的数据进行5折交叉验证对其进行验证。两名经验丰富的操作员的手动校正作为基准真值，使用模拟向量进行数据增强提高了训练的稳健性。该研究将DL方法与手动和标准非人工智能自动MC方法进行了比较，使用最小节段性MBF和MFR评估一致性和诊断准确性。
对于显著冠状动脉疾病（CAD），DL-MC MBF、操作员手动MC MBF的受试者操作特征曲线下面积（AUC）相当（分别为AUC = 0.897、0.892和0.889；p>0.05），标准非人工智能自动MC（AUC = 0.877；p>0.05）且显著高于无MC（AUC = 0.835；p<0.05）。MFR也观察到类似结果。与操作员一致性的95%置信限对于MFR为±0.49ml/g/min（平均差异 = 0.00），对于MBF为±0.24ml/g/min（平均差异 = 0.00）。
DL-MC速度明显更快，但在诊断上与手动MC相当。在接受F-氟吡哒唑PET心肌灌注显像（MPI）的患者中，与标准非人工智能自动MC相比，DL-MC获得的MBF和MFR定量结果与经验丰富的操作员手动校正的结果高度一致。
