FOR 5534: Fast Mapping of Quantitative MR-biosignatures at Ultra-high Magnetic Field

超高场（UHF）强度的磁共振成像（MRI），如7特斯拉，为非侵入性组织表征提供了独特的可能性。然而，目前临床应用尚不多见，临床研究多集中于形态学成像。先进的组织对比，如化学交换饱和转移（CEST）、x核MRI和显微结构成像，已经提供了超越形态学的有价值的信息。这些MRI对比的综合应用将为高洞察力的多光谱MRI提供良好的基础。然而，获得这种核磁共振特征扫描目前受到采集时间长、射频场不均匀性导致的数据质量差、患者运动以及解释大量复杂多光谱数据的难度越来越大等因素的限制。为了充分发挥7T MRI的潜力，我们旨在建立增强形态学成像的“MR生物签名成像”。为此，我们将首先通过开发互补的快速磁共振技术来建立方法基础，以独特的非侵入性表征不同组织，它们的化学成分和微观结构。为了将MR特征转化为病理特异性MR生物特征以进行非侵入性组织表征，我们将使用三个临床研究应用。我们期望MR生物特征一旦建立，将揭示慢性疾病中神经变性和组织变性的早期迹象，并为癌症风险因素提供见解。我们相信，这种磁共振生物签名扫描将比个体对比的总和提供更全面的疾病过程洞察。为了实现这些目标，我们将联合MRI物理学家、工程师、数据科学家和临床医生的努力。为了获得高质量的MR数据，将开发“智能”硬件，将射频（RF）线圈技术与多个接收和发射元件以及集成的多模态RF线圈负载和基于雷达的运动跟踪技术相结合。数据科学将用于识别核磁共振签名扫描的最重要数据特征，并加速数据采集。该研究单位（RU）将建立在埃尔兰根强大的研究环境和基础设施之上。在Friedrich-Alexander Universität Erlangen- n<e:1> rnberg (FAU)，数据科学、机器学习和电气工程领域的杰出研究小组将与Erlangen大学医院（UKER）的研究人员密切合作，即与三个最近成立的研究小组合作，专注于新型MR对比和UHF MRI，并与合作临床研究人员合作。将使用专用的临床7T系统，为结合尖端技术和临床研究提供独特的可能性。

Magnetic resonance imaging (MRI) at ultra-high field (UHF) strengths, such as 7 Tesla, offers unique possibilities for non-invasive tissue characterization. Nevertheless, clinical applications are currently still rare, and present clinical research studies mostly focus on morphological imaging. Advanced tissue contrasts, such as chemical exchange saturation transfer (CEST), X-nuclei MRI and microstructural imaging, have already provided valuable information beyond morphology. The combined application of these MRI contrasts would provide a sound basis for highly insightful multispectral MRI. However, obtaining such an MR-signature scan is currently limited by long acquisition times, poor data quality due to radiofrequency field inhomogeneities, patient motion and the increasing difficulty of interpreting the large amount of complex multispectral data. To fully unleash the potential of 7T MRI, we aim to establish “MR biosignature imaging” augmenting morphological imaging. For this purpose, we will first establish the methodological base by developing complementary fast MR techniques for the unique non-invasive characterization of different tissues, their chemical composition and their microstructure. To turn MR-signatures into pathology-specific MR biosignatures for non-invasive tissue characterization, we will use three clinical research applications. We expect that the MR biosignatures, once established, will reveal early signs of neuro-degeneration, tissue degeneration in chronic diseases and provide insight into cancer risk factors. We are convinced that such an MR biosignature scan would provide a more comprehensive insight into disease processes than the sum of the individual contrasts. To achieve these goals, we will unify the efforts of MRI physicists, engineers, data scientists as well as clinicians. To enable the acquisition of high MR data quality, ‘smart’ hardware will be developed that combines radiofrequency (RF) coil technology with multiple receive and transmit elements as well as integrated multimodal RF coil load- and radar-based motion tracking technology. Data science will be employed to identify the most important data features of the MR-signature scan and to accelerate data acquisition. This research unit (RU) will build on a strong research environment and infrastructure in Erlangen. At the Friedrich-Alexander Universität Erlangen-Nürnberg (FAU), outstanding research groups in the field of data science, machine learning and electrical engineering will contribute by working closely with the researchers of the University Hospital Erlangen (UKER), i.e. with three recently established research groups focusing on novel MR contrasts and UHF MRI, and with collaborating clinical researchers. A dedicated clinical 7T system will be used, offering unique possibilities for combining cutting-edge technological and clinical research.