Image acquisition and analysis tools for magnetic resonance imaging near brain injury

在接受过手术或脑损伤的患者中，在损伤周围区域获得的磁共振成像（MRI）数据可能会出现扭曲或伪影。在核磁共振成像中，射频电磁脉冲被用来激发样本中的信号，而以千赫兹频率切换的电磁场梯度脉冲被用来告诉我们样本中的信号来自何处——从而创建图像。把这两种形式的电磁脉冲结合起来，就形成了一种叫做脉冲序列的工具。在获取MRI数据后，可以重建详细的图像。可以创建新的脉冲序列，以最大限度地减少伪影，并从样品中提取独特的信息。数据采集后，可以开发模型和图像处理工具，将多个脉冲序列获得的信息结合起来，提取特征并隔离单个成像序列获得的数据中不可见的结构。该计划侧重于MR物理研究的两个领域1)调查和开发能够最大限度地减少伪影和失真的脉冲序列；2)开发和测试图像分析工具，这些工具可能有助于将通常难以分析的大脑区域可视化。脉冲序列：我的研究目的是研究新的脉冲序列设计，它可以最大限度地减少技术挑战，并促进在通常被人工制品破坏的环境中改进MRI。这项工作的目标是通过开发新的脉冲序列来解决成像限制，这些脉冲序列结合了加速数据收集策略，使我们能够同时从样品的多个区域获取数据，并优化了信号激励和采集策略，使我们能够最大限度地减少延迟引起的图像失真。图像处理方法：大多数结构分割工具依赖于现有的脑图谱和成像序列，其对比度可能因意外物质的存在而改变。我们将创建改进分割的工具，并能够结合来自不同MRI序列的信息，包括正在开发的新序列。我们将仔细地质疑和测试我们的模型和结果，以减少或消除偏差，并使用我们的分析结果来指导脉冲序列设计。通过这项工作，我打算培养一批多样化的研究生和本科生。我将利用我在科学推广方面的经验，积极鼓励来自不同背景的候选人，以补充被动和部门招聘。有效评估传统上由于伪影而难以成像的样本的能力将扩展这一强大工具的实用性，以包括那些否则无法获得信息丰富的MRI检查的患者。

In patients who have had surgery or brain injuries there may be distortions or artifacts in the magnetic resonance imaging (MRI) data acquired in the region surrounding the injury. In MRI, radio frequency electromagnetic pulses are used to excite a signal from the sample and electromagnetic field gradient pulses, switched at kilohertz frequencies are used to tell us where in the sample that signal is coming from - allowing the creation of an image. Combined, these two forms of electromagnetic pulses create tools called pulse sequences. After the acquisition of MRI data, a detailed image can be reconstructed. Novel pulse sequences can be created to minimize artifacts and extract unique information from the sample. After data acquisition, it is possible to develop models and image processing tools which combine the information obtained by multiple pulse sequences to extract features and isolate structures not visible in the data obtained from a single imaging sequence. This program focuses two areas of MR physics research 1) Investigating and developing pulse sequences which are capable of minimizing artifacts and distortions; and 2) Developing and testing image-analysis tools which may aid in the visualization of regions of the brain that are typically difficult to analyze. Pulse Sequence: My research aims to investigate new pulse sequence designs which can minimize technical challenges and facilitate improved MRI in environments typically corrupted by artifacts. The goal of this work is to address imaging limitations through the development of new pulse sequences that couple accelerated data collection strategies, which allow us to acquire from multiple regions of the sample simultaneously, and optimized signal excitation and acquisition strategies, which allow us to minimize delay-induced image distortions. Image processing methodology: Most structural segmentation tools rely on existing brain atlases as well as imaging sequences whose contrasts may be changed by the presence of unexpected materials. We will create tools which improve segmentation and are capable of combining information from different MRI sequences, including novel ones under development. We will carefully question and test our models and results to reduce or eliminate bias and use the results of our analysis to inform pulse sequence design. Through this work I intend to train a diverse group of graduate and undergraduate students. I will leverage my experience in science outreach to supplement passive and departmental recruitment by actively encourage candidates from a range of backgrounds. The ability to effectively evaluate samples which are traditionally difficult to image due to artifacts will extend the utility of this powerful tool to include patients who would otherwise be unable to access informative MRI exams.