MR Fingerprinting based Quantitative Imaging and Analysis Platform (MRF-QIA) for brain tumors.

摘要 磁共振图像的临床用途在很大程度上是作为一种定性工具而没有内置的标准化， 这需要主观的解读和耗时的分析。重要的是，这些 定性的核磁共振成像方法显示了较差的组织特征，以及较差的中心到中心 中心重复性，极大地限制了它们在临床试验中的使用。提供强大的量化 具有高组织分辨率的成像工具可通过提供可操作的方式直接影响临床护理 提供给最终用户临床医生的信息。例如，准确的肿瘤浸润图的可用性 在胶质母细胞瘤，一种高度侵袭性的脑瘤，可以为新的多部位临床铺平道路 个体化放射治疗和神经外科手术的试验，以改善结果。没有一个是 目前的MRI技术以一种准确和可重复性的方式提供了这种能力。MRF是一种 定量成像扫描可以解决定性MRI的局限性，通过提供 组织特性的可重复性和生理学意义的测量。我们还有 表明利用定量MRF值的基本物理/生理界限 提高了图像分析技术的重现性。MRF与ADVANCE的集成 定量分析可以从根本上解决公认的重复性低的问题 定性的磁共振成像方法，并允许广泛的临床翻译。在这项提案中，我们有 在MRF开发商(CWRU)、图像分析中建立了学术-产业合作伙伴关系 和人工智能专家(UPenn)、脑瘤成像专家(UHCMC)和领先的医疗保健公司 (西门子)，以确保MRF-QIA成功地转化为临床工作流程。我们 将以以下目标实现我们的目标：目标1：建立高通量的MRF扫描和 评估FDA批准的多站点性能；目标2：完全整合MRF-QIA图像 用于脑瘤分析的临床系统分析软件；目标3：临床验证 MRF-QIA在胶质母细胞瘤患者侵袭预测中的应用此项目将添加新的 直接影响最终用户体验和患者护理的临床流程能力：1)FDA MRF产品扫描的批准将允许任何西门子临床站点将其添加到他们的常规患者中 扫描。2)MRF-QIA软件将通过西门子全球数据公司在全球分销 并将用于广泛的临床和多点研究应用。3) GB渗透预测的专门应用将导致新的临床试验计划 神经外科医生的靶向活检、扩大切除和个性化放射治疗 神经肿瘤学家最终为GB患者提供有针对性的治疗计划。

Abstract The clinical utility of MR images is largely as a qualitative tool without in-built standardization, which requires subjective interpretation and time-consuming analysis. Importantly, these qualitative MRI approaches have demonstrated poor tissue characterization, and poor center-to- center reproducibility, greatly limiting their use in clinical trials. Availability of a robust quantitative imaging tool with high tissue discriminability can directly impact clinical care by offering actionable information to end-user clinicians. As an example, availability of accurate tumor infiltration maps in Glioblastomas, a highly aggressive brain tumor, can pave the way for novel multisite clinical trials in personalized radiation therapy and neurosurgery for improved outcomes. None of the current MRI techniques offer this capability in an accurate and reproducible manner. MRF is a quantitative imaging scan that can address the limitations of qualitative MRI by providing reproducible and physiologically meaningful measurements of tissue properties. We have also shown that utilizing the underlying physical/physiological bounds of the quantitative MRF values improves the reproducibility of the image analysis techniques. Integration of MRF and advanced quantitative analytics could fundamentally address the well-recognized low-reproducibility in qualitative MRI approaches and allow broad clinical translation. In this proposal, we have established an academic-industrial partnership among MRF developers (CWRU), image analysis and AI experts (UPenn), Brain tumor imaging experts (UHCMC), and leading healthcare company (Siemens) to ensure successful clinical translation of the MRF-QIA into the clinical workflow. We will achieve our goal with the following aims: Aim 1: Establish a high throughput MRF scan and assess multisite performance for FDA approval; Aim 2: Fully integrate the MRF-QIA image analytics software into the clinical system for brain tumor analysis; Aim 3: Clinical validation of the MRF-QIA application for infiltration prediction in Glioblastoma patients. This project will add new capabilities to the clinical flow directly impacting the end-user experience and patient care: 1) FDA approval of MRF product scan will allow any Siemens clinical site to add it to their routine patient scans. 2) The MRF-QIA software will be distributed globally through Siemens Global Digital Market and will be available for broad clinical and multisite research applications. 3) The specialized application for GB infiltration prediction will lead to new clinical trials for planning targeted biopsy, extended resections, and personalized radiotherapy by neurosurgeons and neuro-oncologists to eventually provide targeted treatment plans for GB patients.