Free-Breathing Perfusion MRI of the Abdomen with Predictive Motion Correction

描述（申请人提供）：随着人口老龄化，肾脏疾病的发病率正在上升。诊断成像在检测和监测肾脏疾病包括肾细胞癌（RCC）中起着关键作用。然而，在成像过程中使用静脉造影剂的风险在患有肾脏疾病和年龄在60岁或以上的个体中增加。本研究的目的是发展无造影剂磁共振成像（MRI）技术在RCC中的应用。MRI程序将包括灌注量化以检测肾脏肿瘤。呼吸运动补偿是腹部MRI的一个主要挑战。像灌注MRI这样的功能性检查需要精确的运动补偿，因为屏气或呼吸门控等替代方法会增加检查时间和患者的压力。本研究的目的是增强ASL MRI在RCC和腹腔灌注MRI中的应用。本提案的具体目的是：1)利用计算机模拟和体内呼吸运动数据识别和评估候选呼吸运动预测（RMP）算法；2)将最佳RMP算法与2D回波平面成像（EPI）和3D梯度和自旋回波（GRASE） MRI相结合，实现腹部自由呼吸快速MRI采集，并利用呼吸运动模体优化技术，比较其在人体志愿者中的表现；3)将最佳2D EPI和3D GRASE RMP技术与脉冲连续ASL (pCASL) MRI序列相结合，评估并优化其在人体志愿者中的表现；4)在人类RCC患者中应用最佳RMP灌注（2D vs 3D pCASL）技术评估肿瘤灌注，并将结果与金标准（CT加对比剂）进行比较。本研究将测试的假设包括：1)肾脏运动可以预测1 s后的未来，误差<2 mm；2)采用回顾性图像后处理的RMP 3D GRASE比采用前瞻性图像切片平移的2D EPI提供更好的运动校正；3)与使用RMP的2D EPI pCASL相比，3D GRASE pCASL具有更好的灌注信噪比（SNR）和更少的运动伪影；4) pCASL联合RMP检测RCC肿瘤边界的准确率为90%。RMP方法将允许自由呼吸和实时跟踪感兴趣的器官，同时最大限度地减少丢弃的图像数据。运动敏感功能MRI序列，如动脉自旋标记、弥散加权成像（DWI）和血氧水平依赖（BOLD），在评估器官健康和肿瘤活性方面也将受益。该技术可以进一步发展用于其他感兴趣的解剖区域或生理运动阻碍准确诊断或治疗的应用（例如，放射治疗）。

DESCRIPTION (provided by applicant): The incidence of renal diseases is growing as the population ages. Diagnostic imaging serves a key role in detecting and monitoring renal diseases including renal cell carcinoma (RCC). Yet the risk of intravenous contrast agents used in imaging procedures rises for individuals with renal disease and aged 60 years or older. The objectives of the proposed research are to develop contrast free magnetic resonance imaging (MRI) techniques for application in RCC. The MRI procedures will include perfusion quantification to detect renal tumors. Respiratory motion compensation is a major challenge in MRI of the abdomen. Functional procedures like perfusion MRI require accurate motion compensation since alternatives like breathholds or respiratory gating increase the length of the exam and the strain on the patient. The objectives of the proposed research are to enhance ASL MRI for application in RCC and abdominal perfusion MRI. The specific aims of this proposal are to: 1) identify and evaluate candidate respiratory motion prediction (RMP) algorithms using computer simulations and in vivo respiratory motion data; 2) integrate the best RMP algorithms with 2D echo planar imaging (EPI) and 3D gradient and spin echo (GRASE) MRI to allow fast free-breathing MRI acquisitions of the abdomen, optimize the techniques using a respiratory motion phantom, and then compare their performances in human volunteers; 3) integrate the best 2D EPI and 3D GRASE RMP techniques with pulsed continuous ASL (pCASL) MRI sequences, and evaluate and optimize its performance in human volunteers; 4) apply the best RMP perfusion (2D vs. 3D pCASL) technique in human subjects with RCC to assess tumor perfusion and compare the results to the gold standard (CT with contrast). The hypotheses that will be tested in this study include: 1) kidney motion can be predicted 1 s into the future with an error of <2 mm; 2) 3D GRASE with RMP using retrospective image post-processing will provide superior motion correction than 2D EPI with prospective image slice translation; 3) 3D GRASE pCASL with RMP will provide superior perfusion signal-to-noise ratio (SNR) and less motion artifact than using 2D EPI pCASL with RMP; 4) pCASL with RMP can detect RCC tumor boundaries with >90% power. The RMP method will permit free-breathing and real-time tracking of the organ of interest while minimizing discarded image data. Motion sensitive functional MRI sequences like arterial spin labeling, diffusion weighted imaging (DWI), and blood oxygenation level dependent (BOLD) that have utility in assessing organ health and tumor activity will also benefit. The technique may be further developed for other anatomical regions of interest or applications where physiological motion impedes accurate diagnosis or therapy (e.g., radiotherapy).