Online MRI Positioning and Real-Time Motion Correction

在线 MRI 定位和实时运动校正

• 批准号: 6734280
• 负责人: Andre Jan Willem van der Kouwe
• 金额: $ 25.51万
• 依托单位: MASSACHUSETTS GENERAL HOSPITAL
• 依托单位国家: 美国
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-09-15 至 2005-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6734280
• 关键词: bioimaging /biomedical imaging; brain; clinical research; human subject; magnetic resonance imaging; method development; neuroimaging

DESCRIPTION (provided by applicant): Automatic on-line slice positioning, on-line slice position correction and prospective real-time motion correction for magnetic resonance imaging of the human brain will be implemented. These technologies will improve the work-flow efficiency of routine clinical evaluations of the brain, and ensure accurate reproducibility of slice positioning over several scanning sessions for the evaluation of progressive illnesses such as tumors and multiple sclerosis. Motion correction will extend the usefulness of MR imaging to traditionally troublesome populations such as children, older subjects and subjects with movement disorders such as Parkinson's disease, where head movements in the scanner frequently preclude successful scanning and confound an accurate diagnosis. Automatic slice positioning obviates the need for the technologist to manually prescribe slices, thereby saving time and improving the reproducibility of scans between scanners, technologists and between sessions in longitudinal studies. The technology is based on a segmented averaged probabilistic atlas and uses a rapid gradient echo (GRE) localizer sequence with two acquisitions with differing contrasts to provide a rapid method for reliably discriminating between the various tissues in the human head and brain. The localizer is collected in approximately forty seconds at the beginning of the scanning session, and to correct for subsequent motion of the subject, two second echo-planar or segmented GRE correction localizers are interleaved between scans. These are referenced to an initial rapid correction localizer to ensure consistent positioning of every scan in the session. Test-retest reliability will be evaluated in a group of healthy volunteers. To correct for within-scan motion, a prospective real-time motion correction scheme is proposed. The scheme requires embedded "octant" navigators of less than 3 ms interleaved between the lines of a conventional 3D sequence such as GRE or 3D EPI. By comparing the navigators with an initial rapidly acquired map of k-space in the vicinity of the navigators, a full rigid body assessment can be made of the position of the object in the scanner at a rate of 50 Hz or more. Corrections to the images can be done off-line, or are made on the scanner in real-time using the rapid feedback mechanism that the Siemens platform provides. The technology will be evaluated in a group of healthy volunteers.

描述（由申请人提供）： 实现人脑磁共振成像自动在线切片定位、在线切片位置校正和前瞻性实时运动校正。这些技术将提高大脑常规临床评估的工作流程效率，并确保多次扫描过程中切片定位的准确再现性，以评估肿瘤和多发性硬化症等进行性疾病。运动校正将把 MR 成像的用途扩展到传统上麻烦的人群，如儿童、老年受试者和患有帕金森病等运动障碍的受试者，扫描仪中的头部运动经常妨碍成功扫描并混淆准确的诊断。自动切片定位消除了技术人员手动指定切片的需要，从而节省了时间并提高了扫描仪、技术人员之间以及纵向研究会话之间扫描的可重复性。该技术基于分段平均概率图谱，并使用快速梯度回波 (GRE) 定位器序列，通过两次不同对比度的采集，提供一种快速方法来可靠地区分人类头部和大脑中的各种组织。在扫描会话开始时大约四十秒内收集定位器，并且为了校正对象的后续运动，在扫描之间交错两个第二回波平面或分段 GRE 校正定位器。这些参考初始快速校正定位器，以确保会话中每次扫描的定位一致。重测可靠性将在一组健康志愿者中进行评估。为了校正扫描内运动，提出了一种前瞻性实时运动校正方案。该方案需要嵌入式“八分圆”导航器，在传统 3D 序列（例如 GRE 或 3D EPI）的行之间交错的时间小于 3 ms。通过将导航器与导航器附近的初始快速获取的 k 空间图进行比较，可以以 50 Hz 或更高的速率对扫描仪中物体的位置进行完整的刚体评估。对图像的校正可以离线完成，也可以使用西门子平台提供的快速反馈机制在扫描仪上实时进行。该技术将在一组健康志愿者中进行评估。