Real-Time Motion Correction and Increased Scan-Session Sucess in Clinical MRI

描述（由申请人提供）：三个fMRI CPT代码（用于临床计费和保险）已被批准并于2007年1月1日生效。临床fMRI将迅速扩展。然而，临床功能磁共振成像经常被过度的患者运动所混淆。研究成果丢失，需要回忆研究对象。此外，在fMRI扫描期间，患者必须充分执行一项行为任务，如果他们感到疲劳或睡着，这可能会受到影响。功能性核磁共振研究通常需要很长时间。这两个问题——患者运动/表现和扫描时间长度——是本提案的重点。改进的方法是目标。三个目标从不同的角度解决问题：目标1：脉冲序列和成像物理用于分离运动伪影的fMRI信号；目标2：实时统计和图像分析，缩短扫描时间，提高抗噪声能力；目标3：使用自适应获取范例、多参数映射和实时临床相关显示来管理扫描会话持续时间。这些目标在针对这两个问题的多方面互动攻击中结合在一起。运动污染的数据要么被实时修复，要么被实时删减，并伴有额外的数据采集。在逐体素的基础上，自适应卡尔曼滤波将调整由慢动作漂移引起的基线变化。更高的接收器带宽可以在几个方面减少运动伪影。我们将探索使用高达1mhz的接收器带宽，具有相应更高的梯度强度。从本质上讲，这是一项相对于扫描仪前置放大器噪声的生理噪声研究。至于扫描时间长度，研究不应超过统计显著结果所需的时间，因为这需要实时统计分析。随机多路行为范式减少了测试时间。此外，所有上述技术都将通过在MRI扫描仪套件中使用增加的计算机能力来实现近实时。最后，我们寻求在患者离开扫描仪之前对扫描会话成功的临床相关确定，允许fMRI用于传统的测试/重新测试范例。这将显著提高功能磁共振成像检查的成功率，从而促进功能磁共振成像作为常规临床工具的采用。研究将在GE 3T扫描仪上进行，该扫描仪于2005年使用国家研究资源中心高端仪器项目提供的资金购买。这项工作将由功能磁共振成像的一个先驱小组进行。功能磁共振成像对脑相关疾病的潜在影响是巨大的。提案中引用了威斯康星医学院正在进行的两项具体研究：可能导致阿尔茨海默病的轻度认知障碍受试者，以及脑肿瘤手术治疗前视觉、运动和语言功能的fMRI测绘。还有更多。

DESCRIPTION (provided by applicant): Three fMRI CPT codes (for clinical billing and insurance) have been approved and went into effect January 1, 2007. Clinical fMRI will expand rapidly. However, clinical fMRI is often confounded by excessive patient motion. Studies are lost and subjects need to be recalled. Moreover, during the fMRI scan, patients must adequately perform a behavioral task, which can suffer if they become fatigued or fall asleep. Functional MRI studies often take too long. These two problems-patient motion/performance and scan-session length-are the focus of this proposal. Improved methodology is the goal. The three aims address the problems from different perspectives: Aim 1: Pulse sequences and imaging physics for separation of fMRI signals from motion artifacts; Aim 2: Real-time statistics and image analysis for shortening scan duration and improvement of noise immunity; Aim 3: Management of scan-session duration with adaptive acquisition paradigms, multi-parameter mapping, and real-time clinically relevant displays. The aims come together in multifaceted interactive attacks on the two problems. Motion-contaminated data will either be fixed in real time or censored in real time accompanied by additional data acquisition. On a voxel-by-voxel basis, adaptive Kalman filtering will adjust baseline variation caused by slow-motion drift. Higher receiver bandwidth can reduce motion artifacts in several ways. We will explore use of receiver bandwidths as high as 1 MHz, with correspondingly higher gradient strengths. At its essence, this is a study of physiological noise relative to scanner preamplifier noise. As for scan-session length, studies should not extend longer than necessary for statistically significant results, which requires real-time statistical analysis. Randomized multiplexed behavioral paradigms reduce test time. In addition, all of the foregoing techniques will be implemented in near-real time through the use of increased computer power in the MRI scanner suite. In the end, we seek a clinically relevant determination of scan-session success before the patient leaves the scanner, allowing fMRI to be used in a conventional test/retest paradigm. This will significantly improve the success rate of fMRI exams and thereby facilitate the adoption of fMRI as a routine clinical tool. Research will be performed on a GE 3T scanner that was purchased in 2005 using funds made available through the High End Instrumentation Program of the National Center for Research Resources. The work will be carried out by one of the pioneering groups in fMRI. The potential impact of fMRI on brain-related disease is enormous. References are made in the proposal to two specific studies underway at the Medical College of Wisconsin: subjects with mild cognitive impairment that may lead to Alzheimer's disease, and fMRI mapping of visual, motor, and speech function prior to surgical treatment of brain tumors. There are many more.