Solving the MRI motion problem with Framewise Integrated Real-Time MRI Monitoring (FIRMM) software

项目摘要/摘要 该SBIR/STTR应用的目标是提供一种准确和非侵入性测量 在结构磁共振成像（MRI）扫描期间患者的头部运动（逐帧集成 实时MRI监测-结构[FIRMM-s]）。因为结构性核磁共振扫描能产生高分辨率 成像并且不使患者暴露于辐射，它已成为非常有价值的诊断工具， 特别是对大脑成像。去年，仅在美国，就有超过800万例脑部核磁共振成像， 估计耗资200 - 300亿美元。不幸的是，脑部核磁共振成像受到以下事实的限制： 扫描可能导致所得到的图像不是最佳的或者甚至是不可用的。估计有20%的大脑 MRI被运动破坏，每年浪费20 - 40亿美元。目前，有两种主要的战略， 战斗头部运动：重复扫描和麻醉，这两个都是不够的。重复扫描， 包括获取额外图像（以确保获取足够的可用图像），增加扫描时间 并且可能导致可用图像太少或不必要的额外图像。麻醉剂，用于 可能移动的患者（如幼儿），存在严重的安全风险，有时 不必要地给药（即，患者可以在没有麻醉的情况下保持静止）。 本授权中提出的基于软件的FIRMM-s解决方案使用MR图像（当它们被收集时）， 在MRI扫描期间计算患者的头部运动的真实的时间。真实的时间运动的可用性 信息将使更多的知情麻醉使用和减少过度扫描，使这些方法 更安全更高效。有了真实的时间运动信息，扫描操作员将准确地知道有多少 已经获取了可用的图像，防止获取过多或过少的额外图像。 此外，为医生提供关于患者运动的定量信息将允许他们做出评估。 关于麻醉的知情决定，防止不必要的镇静。 所提出的解决方案还包含一种对抗头部运动的全新方法：患者生物反馈。 该技术可以将头部运动信息转化为适合年龄的视觉生物反馈， 病人通过向患者提供反馈，该技术可以帮助儿科和成人患者保持更多的 仍然改善图像质量。拟议的研究重点是为FIRMM-s提供概念验证 （第一阶段）和构建和验证FIRMM-s的临床就绪版本（第二阶段）。FIRMM-s设备 为扫描操作员、医生和患者提供真实的时间运动信息， MR扫描更安全、更快、更便宜。

Project Abstract/Summary The goal of this SBIR/STTR application is to deliver a technology that accurately and non-invasively measures a patient’s head motion during a structural magnetic resonance imaging (MRI) scan (Framewise Integrated Real-Time MRI Monitoring -structural [FIRMM-s]). Because structural MRI scanning produces high-resolution images and does not expose patients to radiation, it has become an immensely valuable diagnostic tool, particularly for imaging the brain. Last year, in the United States alone, there were over 8 million brain MRIs, costing an estimated $20-30 billion. Unfortunately, brain MRIs are limited by the fact that head motion during the scan can cause the resulting images to be suboptimal or even unusable. An estimated 20% of all brain MRIs are ruined by motion, wasting $2-4 billion annually. Currently, there are two predominant strategies to combat head motion: repeat scanning and anesthesia, both of which are inadequate. Repeat scanning, which consists of acquiring extra images (to ensure enough usable ones were acquired), increases scanning time and cost, and can result in too few usable images or unnecessary, extra images. Anesthesia, which is given to patients who are likely to move (such as young children), presents a serious safety risk and is sometimes administered unnecessarily (i.e. the patient could hold still without anesthesia). The software-based FIRMM-s solution proposed in this grant uses MR images (as they are being collected) to compute a patient’s head motion in real time during an MRI scan. The availability of real time motion information will enable more informed anesthesia use and reduce excess scanning, making these methods safer and more efficient. Armed with real time motion information, scan operators will know exactly how many usable images have been acquired, preventing the acquisition of too many or too few extra images. Additionally, providing physicians with quantitative information about patient motion will allow them to make an informed decision regarding anesthesia, preventing unnecessary sedation. The proposed solution also contains an entirely new method for combating head motion: patient biofeedback. The technology can translate the head motion information into age-appropriate, visual biofeedback for the patient. By providing feedback to patients, the technology helps both pediatric and adult patients remain more still, improving image quality. The proposed research focuses on delivering proof-of-concept for FIRMM-s (Phase I) and building and validating a clinical-ready version of FIRMM-s (Phase II). The FIRMM-s device provides scan operators, physicians, and patients with real time motion information, with the goal of making MR scans safer, faster, and less expensive.