TimeAware Data Acquisition,RealTime Processing for fMRI

描述(申请人提供)：功能磁共振成像已成为绘制脑功能图的工具，因为它具有高空间和时间分辨率以及非侵入性。此外，它还适用于单个受试者，因此可以在临床领域发挥作用。这些特性对临床人群、研究、治疗和疾病监测具有吸引力。然而，与正常志愿者相比，临床人群更有可能搬家，更有可能无法完成任务，对长时间成像过程的容忍度较低。因此，这些受试者可能会产生比正常更少的数据和更低的质量，从而更难检测到大胆对比功能磁共振固有的低信号变化。此外，合格的患者比正常情况下更难获得，因此确保获得有用的数据至关重要。从结果中及时反馈对于治疗计划也是必要的。因此，我们已经着手开发一种临床上有能力的功能磁共振成像系统，它比现有系统更灵活和更高效。该系统的目标是在存在多个并发效应的情况下提供对激活的时间效率估计，跟踪导致混杂效应(运动、语音、心脏搏动、呼吸等)的过程，并使处理适应于受试者的表现而不是预期的任务。该系统的主要组成部分，如时间感知架构、生理数据(语音、心脏、呼吸、皮肤电导)的测量和过滤以及实时多元线性回归都已完成。我们将通过结合运动检测和校正、生理噪声建模和开发实时语音检测来扩展系统，以用于范式，其中范式传递的时间和数据的处理都由受试者的反应驱动。

DESCRIPTION (provided by applicant): FMRI has become a tool for mapping brain function because of its high spatial and temporal resolution and noninvasiveness. In addition, it is applicable to single subjects and thus could play a role in the clinical realm. These properties are appealing for use in clinical populations, for research, treatment and disease monitoring. However, in contrast to normal volunteers, clinical populations are more likely to move, more likely to fail to conform to the task and are less tolerant to long imaging sessions. Thus, these subjects may generate less data and of lower quality than normal making it more difficult to detect the low signal changes inherent in BOLD contrast fMRI. In addition, qualified patients are more difficult to come by than normal making it critical to insure that useful data has been acquired. Timely feedback from results is also necessary for treatment planning. Thus, we have set out to develop a clinical competent fMRI system that is more flexible and more efficient that existing systems. The goal of the system is to provide time efficient estimation of activation in the presence of multiple concurrent effects, track processes that lead to confounding effects (motion, speech, cardiac pulsatility, respiration, etc) and adapt the processing to the subject's performance rather than the expected task. Major components of this system such as the timeaware architecture, measurement and filtering of physiologic data (speech, cardiac, respiration, skin conductance) and real-time multiple linear regression have been completed. We will extend the system by incorporatingdetection of and correction for motion, modeling of physiologic noise and development of real-time speech detection for use in a paradigm where both the timing of paradigm delivery and processing of the data is driven by the subject's response.