A Quality Control Program for Quantitative MRI Data Acquisition and Analysis in Multi-site Clinical Trials

多站点临床试验中定量 MRI 数据采集和分析的质量控制程序

• 批准号: 10007270
• 负责人: Todor Karaulanov
• 金额: $ 84.37万
• 依托单位: CALIBERMRI, INC.
• 依托单位国家: 美国
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-09-20 至 2021-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10007270
• 关键词: Algorithms; American College of Radiology Imaging Network; Artificial Intelligence; Breast; Breast Magnetic Resonance Imaging; Breast conservation; Clinical Management; Clinical Trials; Complement; Computer software; Contrast Media; Data; Data Analyses; Detection; Development; Diagnostic; Diagnostic Neoplasm Staging; Diffusion; Digital Imaging and Communications in Medicine; Disease; Disease regression; Educational process of instructing; Elements; Ensure; Evaluation; Evidence based treatment; Fatty acid glycerol esters; Genes; Goals; Head; Hour; Image; Image Analysis; In complete remission; Individual; Institutes; Lead; Location; Magnetic Resonance Imaging; Magnetism; Malignant Neoplasms; Manufacturer Name; Maps; Measurement; Measures; Methodology; Methods; Modification; Monitor; Multi-Institutional Clinical Trial; Noise; Operative Surgical Procedures; Outcome; Pathologic; Patient-Focused Outcomes; Patients; Performance; Phase; Physicians; Pilot Projects; Precision Medicine Initiative; Predisposition; Primary Neoplasm; Procedures; Process; Production; Protocol Compliance; Protocols documentation; Protons; Quality Control; Reading; Relaxation; Reporting; Reproducibility; Scanning; Signal Transduction; Site; Standardization; System; Technology; Testing; Time; Tissues; Treatment Efficacy; Tumor Volume; Uncertainty; Variant; Vendor; Work; base; breast imaging; chemotherapy; cloud based; computer aided detection; data acquisition; deep learning; human tissue; imaging biomarker; improved; in vivo; interest; malignant breast neoplasm; nervous system disorder; non-invasive monitor; novel; operation; precision medicine; predictive marker; programs; protocol development; quality assurance; quantitative imaging; side effect; tool; treatment response; tumor

Project Summary/Abstract There has been a shift to greater use of neadjuvant chemotherapy (NACT) in the clinical management of cancer. Often used in breast cancer to provide systemic treatment for possible or apparent metastatic disease, NACT also reduces the primary tumor size and facilitates breast conservation. Quantitative magnetic resonance imaging (qMRI) is an effective diagnostic for monitoring tumor regression. There are many methodologies being developed for monitoring cancer in vivo, including contrast agent washout, volumetry, and more recently correlating apparent diffusion coefficients and proton relaxation times to tumor staging. Such imaging biomarkers are predictive of the pathological complete response (pCR) outcome prior to surgery. This enables physicians to provide “evidence based” treatment modifications to reduce deleterious side effects and improve patient outcomes. To obtain quantitative data from MR images, the variations that occur in imaging as a result of hardware and software differences between MRI manufacturers and day-to-day operations must be quantified. In MRI, the data depends on the protocols used to acquire the image, resulting in inconsistencies if even one parameter in the protocol changes. The goal of the proposed work is to produce and develop and disseminate a ground truth physical standard and quality control program for comparison of MRI scanners in the multisite I-SPY2 (ACRIN 6698) trials monitoring pre-operative chemotherapy treatments in the reduction of tumor size. The physical standard is an anthropomorphic breast phantom capable of fitting into all commercial breast coils. It contains novel human tissue mimics and geometric standards with values traceable to the National Institute of Standards and Technology (NIST). Analysis software capable of protocol compliance checks, rapid identification of the regions of interest, and comparison to ground truth values as measured by NIST will be developed to complement the physical standard. This will enable qualification of sites and quantification of error resulting from scanner performance alone. Given the wide breadth of parameter space that MRI protocols encompass, part of the proposed effort will include the development of protocols specific to image acquisition of tumor volumes and for measuring the apparent diffusion coefficient within the tumor volumes as compared to surrounding healthy tissue.

项目总结/摘要 新辅助化疗（NACT）的临床管理已经转向更多地使用新辅助化疗（NACT）。 癌通常用于乳腺癌，为可能或明显的转移性疾病提供全身治疗， NACT还可以缩小原发肿瘤的大小，促进乳腺保护。定量磁 共振成像（qMRI）是监测肿瘤消退的有效诊断。有很多 正在开发的用于体内监测癌症的方法，包括造影剂洗脱，体积测定， 以及最近将表观扩散系数和质子弛豫时间与肿瘤分期相关联。等 成像生物标志物可预测手术前的病理完全反应（pCR）结果。这 使医生能够提供“基于证据”的治疗修改，以减少有害的副作用， 改善患者结果。 为了从MR图像中获得定量数据，由于硬件和 必须量化MRI制造商和日常操作之间的软件差异。在MRI中， 数据取决于用于获取图像的协议，如果其中的一个参数 协议就变了拟议工作的目标是产生、发展和传播一个基础， 多站点I-SPY 2中MRI扫描仪比较的真实物理标准和质量控制程序 （ACRIN 6698）试验监测术前化疗治疗在缩小肿瘤大小方面的作用。 物理标准是一个拟人乳房体模，能够适应所有商业乳房线圈。 它包含新颖的人体组织模拟物和几何标准，其值可追溯到国家研究所 标准与技术（NIST）分析软件能够进行方案合规性检查，快速 识别感兴趣的区域，并与NIST测量的地面真实值进行比较， 以补充物理标准。这将使研究中心的资格和量化 仅由扫描仪性能导致的错误。考虑到MRI协议的参数空间的广泛性， 包括，部分拟议的努力将包括制定具体的协议，以图像采集 并用于测量肿瘤体积内的表观扩散系数， 周围的健康组织。