School of Biomedical Engineering, Faculty of Engineering and Information Technologies, The University of Sydney, Sydney, Australia.
Geneva University Hospitals, Division of Nuclear Medicine & Molecular Imaging, CH-1205 Geneva, Switzerland.
Phys Med Biol. 2023 Jul 31;68(16). doi: 10.1088/1361-6560/acdf3a.
. Cerebral CT perfusion (CTP) imaging is most commonly used to diagnose acute ischaemic stroke and support treatment decisions. Shortening CTP scan duration is desirable to reduce the accumulated radiation dose and the risk of patient head movement. In this study, we present a novel application of a stochastic adversarial video prediction approach to reduce CTP imaging acquisition time.. A variational autoencoder and generative adversarial network (VAE-GAN) were implemented in a recurrent framework in three scenarios: to predict the last 8 (24 s), 13 (31.5 s) and 18 (39 s) image frames of the CTP acquisition from the first 25 (36 s), 20 (28.5 s) and 15 (21 s) acquired frames, respectively. The model was trained using 65 stroke cases and tested on 10 unseen cases. Predicted frames were assessed against ground-truth in terms of image quality and haemodynamic maps, bolus shape characteristics and volumetric analysis of lesions.. In all three prediction scenarios, the mean percentage error between the area, full-width-at-half-maximum and maximum enhancement of the predicted and ground-truth bolus curve was less than 4 ± 4%. The best peak signal-to-noise ratio and structural similarity of predicted haemodynamic maps was obtained for cerebral blood volume followed (in order) by cerebral blood flow, mean transit time and time to peak. For the 3 prediction scenarios, average volumetric error of the lesion was overestimated by 7%-15%, 11%-28% and 7%-22% for the infarct, penumbra and hypo-perfused regions, respectively, and the corresponding spatial agreement for these regions was 67%-76%, 76%-86% and 83%-92%.. This study suggests that a recurrent VAE-GAN could potentially be used to predict a portion of CTP frames from truncated acquisitions, preserving the majority of clinical content in the images, and potentially reducing the scan duration and radiation dose simultaneously by 65% and 54.5%, respectively.
. 脑 CT 灌注（CTP）成像最常用于诊断急性缺血性中风并支持治疗决策。缩短 CTP 扫描时间是理想的，可以降低累积辐射剂量和患者头部运动的风险。在这项研究中，我们提出了一种新颖的随机对抗视频预测方法的应用，以减少 CTP 成像采集时间。在三个场景中，实现了变分自编码器和生成对抗网络（VAE-GAN）的递归框架：从采集的前 25（36 秒）、20（28.5 秒）和 15（21 秒）个采集帧中分别预测 CTP 采集的最后 8（24 秒）、13（31.5 秒）和 18（39 秒）个图像帧。该模型使用 65 个中风病例进行训练，并在 10 个未见过的病例上进行测试。预测的帧在图像质量和血流动力学图、对比剂形状特征和病变体积分析方面与真实值进行了评估。在所有三个预测场景中，预测和真实值的 bolus 曲线的面积、半最大值全宽和最大增强之间的平均百分比误差均小于 4±4%。在预测的血流动力学图中，获得了最佳的峰值信噪比和结构相似性，其次是脑血容量，然后是脑血流量、平均通过时间和峰值时间。对于 3 个预测场景，对于梗死、半影和低灌注区域，病变的平均体积误差分别高估了 7%-15%、11%-28%和 7%-22%，这些区域的相应空间一致性分别为 67%-76%、76%-86%和 83%-92%。这项研究表明，递归 VAE-GAN 可能可以用于从截断采集预测 CTP 帧的一部分，在图像中保留大部分临床内容，并通过 65%和 54.5%分别同时潜在地减少扫描时间和辐射剂量。
