Yalcinkaya Dilek M, Youssef Khalid, Heydari Bobak, Wei Janet, Merz Noel Bairey, Judd Robert, Dharmakumar Rohan, Simonetti Orlando P, Weinsaft Jonathan W, Raman Subha V, Sharif Behzad
Laboratory for Translational Imaging of Microcirculation, Indiana University School of Medicine, Indianapolis, IN, USA.
Elmore Family School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN, USA.
ArXiv. 2024 Aug 9:arXiv:2408.04805v1.
Fully automatic analysis of myocardial perfusion MRI datasets enables rapid and objective reporting of stress/rest studies in patients with suspected ischemic heart disease. Developing deep learning techniques that can analyze multi-center datasets despite limited training data and variations in software (pulse sequence) and hardware (scanner vendor) is an ongoing challenge.
Datasets from 3 medical centers acquired at 3T (n = 150 subjects; 21,150 first-pass images) were included: an internal dataset (inD; n = 95) and two external datasets (exDs; n = 55) used for evaluating the robustness of the trained deep neural network (DNN) models against differences in pulse sequence (exD-1) and scanner vendor (exD-2). A subset of inD (n = 85) was used for training/validation of a pool of DNNs for segmentation, all using the same spatiotemporal U-Net architecture and hyperparameters but with different parameter initializations. We employed a space-time sliding-patch analysis approach that automatically yields a pixel-wise "uncertainty map" as a byproduct of the segmentation process. In our approach, dubbed Data Adaptive Uncertainty-Guided Space-time (DAUGS) analysis, a given test case is segmented by all members of the DNN pool and the resulting uncertainty maps are leveraged to automatically select the "best" one among the pool of solutions. For comparison, we also trained a DNN using the established approach with the same settings (hyperparameters, data augmentation, etc.).
The proposed DAUGS analysis approach performed similarly to the established approach on the internal dataset (Dice score for the testing subset of inD: 0.896 ± 0.050 vs. 0.890 ± 0.049; p = n.s.) whereas it significantly outperformed on the external datasets (Dice for exD-1: 0.885 ± 0.040 vs. 0.849 ± 0.065, p < 0.005; Dice for exD-2: 0.811 ± 0.070 vs. 0.728 ± 0.149, p < 0.005). Moreover, the number of image series with "failed" segmentation (defined as having myocardial contours that include bloodpool or are noncontiguous in ≥1 segment) was significantly lower for the proposed vs. the established approach (4.3% vs. 17.1%, p < 0.0005).
The proposed DAUGS analysis approach has the potential to improve the robustness of deep learning methods for segmentation of multi-center stress perfusion datasets with variations in the choice of pulse sequence, site location or scanner vendor.
对心肌灌注磁共振成像数据集进行全自动分析，能够对疑似缺血性心脏病患者的负荷/静息研究进行快速且客观的报告。尽管训练数据有限，且软件（脉冲序列）和硬件（扫描仪供应商）存在差异，但开发能够分析多中心数据集的深度学习技术仍是一项持续的挑战。
纳入了来自3个医学中心在3T场强下采集的数据集（n = 150名受试者；21,150幅首过图像）：一个内部数据集（inD；n = 95）和两个外部数据集（exD；n = 55），用于评估训练后的深度神经网络（DNN）模型在脉冲序列差异（exD - 1）和扫描仪供应商差异（exD - 2）情况下的稳健性。inD的一个子集（n = 85）用于训练/验证一组用于分割的DNN，所有DNN均使用相同的时空U - Net架构和超参数，但参数初始化不同。我们采用了一种时空滑动补丁分析方法，该方法会自动生成一个逐像素的“不确定性图”作为分割过程的副产品。在我们称为数据自适应不确定性引导时空（DAUGS）分析的方法中，给定的测试病例由DNN组中的所有成员进行分割，并利用生成的不确定性图自动从一组解决方案中选择“最佳”方案。为作比较，我们还使用相同设置（超参数、数据增强等）的既定方法训练了一个DNN。
所提出的DAUGS分析方法在内部数据集上的表现与既定方法相似（inD测试子集的Dice分数：0.896 ± 0.050对0.890 ± 0.049；p = 无显著差异），而在外部数据集上则显著优于既定方法（exD - 1的Dice分数：0.885 ± 0.040对␣0.849 ± 0.065，p < 0.005；exD - 2的Dice分数：0.811 ± 0.070对0.728 ± 0.149，p < 0.005）。此外，与既定方法相比，所提出方法中分割“失败”（定义为心肌轮廓包含血池或在≥1个节段中不连续）的图像序列数量显著更低（4.3%对17.1%，p < 0.0005）。
所提出的DAUGS分析方法有可能提高深度学习方法对多中心负荷灌注数据集进行分割的稳健性，这些数据集在脉冲序列选择、站点位置或扫描仪供应商方面存在差异。
