Quantitative Magnetic Resonance Relaxometry and Diffusometry

磁共振成像（MRI）是一种灵活而强大的非侵入性诊断成像技术，可用于临床医学。MRI作为一种很大程度上定性的工具已经建立，主要用于通过绘制氢的空间位置来可视化生物系统的内部结构。磁共振信号寿命（弛豫时间）和分子扩散通常定性地体现在磁共振图像对比中。这些参数的定量分析将揭示分子尺度信息，因为磁共振对一系列物理/化学特征很敏感，包括分子结构、运动、大小和相互作用，通过弛豫时间（松弛测量）和分子扩散（扩散测量）的测量。定量磁共振成像的潜力已得到认可。但是，要充分发挥这一潜力，仍然需要在方法上取得重大进展。******我们的目标是用传统的核磁共振仪器定量地绘制弛豫和扩散相关性，实现超出传统方法限制的空间分辨率，以及利用低成本手持式核磁共振扫描仪进行现场测量的组织结构定量分析。我们将在这两种情况下利用基于人工神经网络（深度学习）的数据处理来克服这些测量中固有的挑战，并能够有效地恢复潜在组织的数量和性质。******定量和稳健的松弛法和扩散法将赋予磁共振能力，并将其作为一种分析方法提升到一个新的水平，并显著扩大其在疾病检测和量化方面的能力。******由于这项工作的跨学科性质，涉及生物学和计算机科学以及物理学，这项研究显示出与这些领域和其他领域的实际问题有重大关联的希望。与温莎大学和其他地方的同事合作，将使这些途径得以探索。

Magnetic resonance imaging (MRI) is well known as a flexible and powerful non-invasive diagnostic imaging technique available to clinical medicine. MRI as a largely qualitative tool is well established, principally used to visualize the internal structure of biological systems through mapping the spatial position of hydrogen. The magnetic resonance signal lifetimes (relaxation times) and molecular diffusion usually manifest qualitatively in the MR image contrast. Quantitative analysis of these parameters will reveal molecular scale information, since magnetic resonance is sensitive to a range of physical/chemical characteristics including molecular structure, motion, size and interactions, through the measurements of relaxation times (relaxometry) and molecular diffusion (diffusometry). The potential for quantitative MR has been recognized. However, significant methodological developments are still required for this potential to be fully realized. ****** Our goal is to quantitatively map the relaxation and diffusion correlations with conventional MRI instruments, achieving a spatial resolution beyond the limits set by traditional methods, as well as quantitative analysis of tissue structure with spot measurements utilizing a low cost handheld MR scanner. We will exploit artificial neural network (deep learning) based data processing in both scenarios to overcome the challenges inherent in these measurements and to enable efficient recovery of underlying tissue quantity and property.****** Quantitative and robust relaxometry and diffusometry will empower magnetic resonance and bring it to a new level as an analytical methodology and significantly enlarge its capacity in disease detection and quantification.****** Because of the inter-disciplinary nature of the work, involving biology and computer science as well as physics, the research shows promise of significant relevance to practical problems in these and other fields. Collaboration with colleagues at the University of Windsor and elsewhere will enable these avenues to be explored.