Novel technique for breast cancer screening and evaluation using ultra low field MRI

项目总结/摘要 乳房X线摄影是最常用的基于成像的筛查工具，因为它是可访问的和具有成本效益的。 然而，乳房X光检查有其局限性。乳腺X线摄影的灵敏度只有70- 90%， 阳性预测值范围从20- 80%。此外，乳房X光检查的准确性降低 在致密的乳房中。2015年，只有65.3%的美国女性接受了推荐的乳房X光检查。 女性选择不做乳房X光检查有几个原因。个人因素，如不能容忍 乳房X线摄影成像所需的压缩，对乳房X线摄影中使用的辐射的担忧以及焦虑 假阳性是乳腺X光筛查的障碍目前可用的基于MRI的方法可以 克服其中的一些问题。磁共振成像不使用压缩，不涉及辐射。先生 成像是有效的，因为软组织中的差异可以可视化，而不会使致密组织模糊。 组织中实际上，T1和T2弛豫时间和表观扩散的定量MRI测量 ADC值可以区分正常组织和肿瘤组织。 我们建议使用定量磁共振成像（MRI）的乳腺癌在超低磁 磁场强度（6.5 mT）。具体来说，我们将使用一种称为T1 rho分散的定量MRI技术， 在ULF处生成脂肪、纤维腺体和肿瘤组织的图。T1 rho色散受RF沉积的限制 临床场强（1.5 T和3 T）下的安全性问题;然而，由于这些安全性问题与 磁场强度的平方，这在6.5 mT时不是问题。我们通过两个项目来解决这个问题 互补的目标：首先，我们将收集T1 rho分散体数据在体内使用局部光谱，以评估 这一想法的可行性（目标1）。这一目标需要构建用于ULF的乳房线圈。然后，我们将开发一个 基于SSFP序列结合AI重建的T1 rho弥散成像实验（目的2）。 为此，我们将构建一个乳房体模或参考对象，以模仿超低频下的乳房组织特性。 该项目将评估使用T1 rho SSFP序列快速区分脂肪、纤维腺体和脂肪组织的可行性。 和肿瘤组织，并将比较组织的T1 ρ分散分类（例如，健康、良性、恶性） 活检组织的病理报告 超低场T1 rho弥散MRI可以使基于MRI的筛查工具广泛采用， 乳腺癌没有放射或造影剂的风险，并且在致密组织中的确定性增加。一个超- 低场MRI系统将比传统MRI更便宜，大大降低了选址要求，如 乳房X光检查中心，并用作筛查工具，将使高风险人群的频繁成像成为可能 并且可以增加那些目前选择退出乳房X线照相术的人的依从性。

Project Summary/Abstract Mammography is the most used imaging-based screening tool, because it is accessible and cost-effective. However, mammography has its limitations. The sensitivity of mammography is only 70-90%, and it has a positive predictive value that ranges widely from 20-80%. Additionally, the accuracy of mammography decreases in dense breasts. In 2015 only 65.3% of women in the United States had had their recommended mammogram. Women opt-out of mammography for several reasons. Individual factors such as inability to tolerate the compression needed for mammographic imaging, concerns about radiation used in mammograms and anxiety about false positives are all barriers to screening mammography. Currently available MRI-based methods can overcome some of these issues. MR imaging does not use compression and does not involve radiation. MR imaging is effective because differences in soft tissues can be visualized without obfuscations from dense tissues. Indeed, quantitative MRI measurements of T1 and T2 relaxation times and Apparent Diffusion Coefficient (ADC) can distinguish healthy and tumor tissues. We propose the use of quantitative magnetic resonance imaging (MRI) for breast cancer at ultra-low magnetic field (ULF) strength (6.5 mT). Specifically, we will use a quantitative MRI technique called T1rho dispersion to generate a map of fat, fibroglandular and tumor tissues at ULF. T1rho dispersion is limited by RF deposition safety concerns at clinical field strengths (1.5 T and 3 T); however, because these safety concerns scale with the square of magnetic field strength, that is not an issue at 6.5 mT. We address the project through two complimentary aims: First, we will collect T1rho dispersion data in vivo using localized spectroscopy to assess the feasibility of this idea (Aim 1). This aim requires constructing breast coils for ULF. Then, we will develop a T1rho dispersion imaging experiment based on an SSFP sequence combined with AI reconstruction (Aim 2). For this aim, we will construct a breast phantom or reference object to mimic the breast tissue properties at ULF. This project will assess the feasibility of using the T1rho SSFP sequence to rapidly distinguish fat, fibroglandular and tumor tissues and will compare T1rho dispersion categorization of tissues (e.g., healthy, benign, malignant) with pathology reports of biopsied tissue. Ultra-low field T1rho dispersion MRI may enable the widespread adoption of an MRI-based screening tool for breast cancer without risk of radiation or contrast agents and with increased certainty in dense tissues. An ultra- low field MRI system would be cheaper than conventional MRI, has greatly reduced siting requirements, like mammography centers, and used as a screening tool, would enable frequent imaging in high-risk populations and may increase compliance for those who currently opt-out of mammography.