Inter-vendor standardisation and quality assurance of quantitative cardiovascular magnetic resonance (CMR) T1-mapping for clinical practice and trials

定量CMR（qCMR）T1标测被欧洲心脏病学会评为评估心脏病的最具创新性的技术之一。T1标测提供了身体组织的像素级特征，超出了传统MRI的范围，并已被证明可以检测亚临床心脏和肝脏疾病。每种组织类型都有一个正常的T1范围，偏离该范围可能表明疾病。然而，有许多T1映射技术，每种技术在不同供应商的不同扫描仪上都有自己的正常范围。这种标准化的缺乏限制了中心间的可比性和临床应用的更广泛推广。我们已经成功开发了一种心脏T1标测技术-Shortened Modified Look-Tumor Inversion Recovery（ShMOLLI）-目前在5个国际指南中，有超过60篇论文，约6000次引用和5个奖项。它具有任何单一T1标测方法的最大临床证据基础，是英国生物库中指定的T1标测技术（到2021/2022年将有100，000个数据集）。我们最近还发布了我们为临床试验绘制标准化的独特方法（2021年），并获得了4个行业合作伙伴（包括GE）的支持信，全球有超过100个感兴趣的CMR研究中心。我们的方法目前在单一供应商平台（西门子）上运行，但现在已准备好进行供应商间标准化。考虑到供应商之间在技术上的巨大差异，实现这一目标是一项雄心勃勃的壮举;这需要行业合作伙伴的大量购买和投资，以便在供应商专有系统上提供特定于供应商的方法实现。在解决这一未满足的临床需求后，我们希望通过使用共同的规范和指南，将qCMR映射转化为NHS的广泛使用。这也将允许建立一个独立于供应商的国家成像数据库，与英国生物银行兼容，通过大数据和人工智能解决重要的临床问题。本研究将依赖于最近购买的GE MR扫描仪作为平台，以实现供应商间标准化计划。这个拟议的项目与MRC在几个领域的职权范围相一致：诊断和数据分析的转化影响，使未来的国家成像数据库与英国生物银行大数据相一致，进一步应用人工智能和机器学习，实现精准医学。该项目取决于生物医学成像，MR物理学，医生和行业合作伙伴之间的跨学科技能和工作方式，并将使研究人员能够在卫生和物理科学部门之间流动。

Quantitative CMR (qCMR) T1-mapping is named one of the most innovative technologies to evaluate heart disease by the European Society of Cardiology. T1-mapping provides pixel-wise characterisation of body tissue, beyond what conventional MRI offers, and has been shown to detect subclinical cardiac and liver disease. Each tissue type has a normal T1 range, deviation from which may indicate disease. However, there are many T1- mapping techniques, each with their own normal range on different scanners by different vendors. This lack of standardisation limits inter-centre comparability and wider roll-out for clinical use. We have successfully developed a cardiac T1-mapping technique - the Shortened Modified Look-Locker inversion recovery (ShMOLLI) - now in 5 international guidelines, with >60 papers, ~6000 citations and 5 awards. It has the largest clinical evidence base for any single T1-mapping method, and is the designated T1-mapping technique in UK Biobank (100,000 datasets by 2021/2022). We have also recently published our unique approach to mapping standardisation for clinical trials (2021), and had garnered support letters from 4 industry partners (GE included), with >100 interested CMR sites worldwide. Our method currently operates on a single vendor platform (Siemens), but now ready for inter-vendor standardisation. Given substantial inter-vendor differences in technologies, achieving this is an ambitious feat; this requires significant buy-in and investment from industry partners, to provide vendor-specific method implementation on vendor proprietary systems. Upon solving this unmet clinical need, we expect to translate qCMR mapping into widespread use in the NHS, through the use of common norms and guidelines. This would also allow the building of a vendor-independent national imaging database, compatible with UK Biobank, to address important clinical questions through big data and artificial intelligence. This study will rely on recently purchased GE MR scanner as a platform to enable inter-vendor standardisation programme. This proposed project is aligned to MRC's remit in several areas: the translational impact of diagnostic and data analytics, to enable a future national imaging database that is aligned with the UK Biobank big data, for further applications of artificial intelligence and machine learning, towards precision medicine. The project hinges on interdisciplinary skills and ways of working, between biomedical imaging, MR physics, physicians and industry partners, and will enable researcher mobility between sectors of health and the physical sciences.