Prostate-cancer Imaging by Combined Ultrasound and Magnetic-resonance Parameters

结合超声和磁共振参数的前列腺癌成像

• 批准号: 8677765
• 负责人: Ernest Joseph Feleppa
• 金额: $ 55.35万
• 依托单位: RIVERSIDE RESEARCH INSTITUTE
• 依托单位国家: 美国
• 财政年份: 2010
• 资助国家: 美国
• 起止时间: 2010-08-09 至 2016-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/8677765
• 关键词: Ablation; Achievement; Acoustics; Address; Adverse effects; Area; Atlas of Cancer Mortality in the United States; Biological Neural Networks; Biopsy; Biopsy Specimen; Bladder; Blood; Blood Vessels; Chemicals; Classification; Clinical; Cryosurgery; Data; Data Analyses; Databases; Developed Countries; Development; Diagnosis; Diffusion; Diffusion Magnetic Resonance Imaging; Disease; Disease Progression; Disease regression; Drug Formulations; Engineering; Frequencies; Generations; Gland; Goals; Health; Histocompatibility Testing; Histology; Image; Israel; Lead; Machine Learning; Magnetic Resonance; Magnetic Resonance Imaging; Malignant Neoplasms; Malignant neoplasm of prostate; Mechanics; Medical center; Metabolic; Methods; Microscopic; Modality; Monitor; Morphologic artifacts; NIH Program Announcements; Neoplasm Metastasis; Nerve; Operating Rooms; Operative Surgical Procedures; Patients; Performance; Perfusion; Process; Property; Prostate; Prostate Cancer therapy; Prostate-Specific Antigen; Prostatectomy; Quantitative Evaluations; ROC Curve; Radio; Radiosurgery; Rectum; Research; Research Institute; Resolution; Risk; Sensitivity and Specificity; Serum; Signal Transduction; Specimen; Spectrum Analysis; Staging; Structure of base of prostate; Surface; System; Technical Expertise; Techniques; Technology; Time; Tissues; Ultrasonography; United States; Universities; Urethra; Ursidae Family; base; blind; cancer diagnosis; cancer imaging; cancer therapy; computer aided detection; computerized; cost; experience; high risk; image registration; imaging modality; improved; instrument; instrumentation; interest; lymph nodes; novel; outcome forecast; prototype; quantitative ultrasound; quantum; response; standard care; tool; treatment planning; tumor

DESCRIPTION (provided by applicant): Riverside Research Institute, Beth Israel Deaconess Medical Center, Rutgers University, and the General Electric Corporation propose to undertake an Academic-Industrial Partnership study to improve prostate- cancer (PCA) imaging markedly and thereby improve prostate-biopsy guidance to detect PCa; enhance monitoring, surveillance, and treatment of PCa; and enable planning and execution of focal PCa therapy. The project is highly significant because it addresses a major health problem in the United States and other developed countries. The project will overcome the current inability of established clinical-imaging method to image PCa reliably by combining the capabilities of advanced ultrasound (US) and magnetic-resonance (MR) techniques in a clinically effective manner. The proposed approach will exploit the sensitivity of US to mechanical properties of tissue on a microscopic scale with the sensitivity of MR to chemical constituents of tissue and its ability to sense blood distribution. Each of these modalities senses different and independent properties of tissue and has shown encouraging potential for improved imaging of PCa when used alone; combining parameters derived from each modality can provide far superior sensitivity and specificity for PCa. We will combine US and MR parameters using advanced classifiers such as artificial neural networks and support-vector machines. These classifiers already have produced ROC-curve areas of 0.91 for advanced US methods, and the MR methods have demonstrated equivalent ROC-curve areas in many studies. We will embody the combined capabilities in specifications for a prototype imaging system that can generate prostate tissue-typing images (TTIs) in real-time for targeting biopsies or planning treatment in the operating room or in an off-line setting. The latest Logiq E9 instrument currently being produced by GE already has a capability for fusing previously obtained MR images with US images in real time, which provides an existing framework for combining US and MR parameters and generating real-time TTIs. Successfully generating reliable prostate TTIs based on combined US and MR parameters will represent a quantum advance in PCa management by enabling significant improvements in the diagnosis and treatment of PCa.

描述(由申请人提供)：河滨研究所、贝丝以色列女执事医学中心、罗格斯大学和通用电气公司提议开展一项学术-产业伙伴关系研究，以显著改善前列腺癌(PCA)成像，从而改善前列腺活检指导，以检测前列腺癌；加强对前列腺癌的监测、监测和治疗；并使重点前列腺癌治疗的规划和实施成为可能。该项目意义重大，因为它解决了美国和其他发达国家的一个主要健康问题。该项目将通过将先进的超声(US)和磁共振(MR)技术的能力以临床有效的方式结合起来，克服目前已建立的临床成像方法无法可靠地成像PCA的问题。该方法将在微观尺度上利用US对组织机械特性的敏感性，以及MR对组织化学成分的敏感性和感知血液分布的能力。这些模式中的每一种都可以感知不同且独立的组织属性，并且在单独使用时显示出改善PCA成像的令人鼓舞的潜力；结合来自每种模式的参数可以为PCA提供更高的灵敏度和特异度。我们将使用人工神经网络和支持向量机等高级分类器来组合US和MR参数。这些分类器已经为先进的US方法产生了0.91的ROC曲线区域，而MR方法在许多研究中证明了等同的ROC曲线区域。我们将在原型成像系统的规格中体现组合功能，该系统可以实时生成前列腺组织分型图像(TTI)，用于在手术室或脱机环境中进行靶向活检或计划治疗。通用电气目前正在生产的最新Logiq E9仪器已经具备了将先前获得的MR图像与US图像实时融合的能力，这为组合US和MR参数并生成实时TTI提供了现有框架。成功地基于US和MR参数组合生成可靠的前列腺TTI将使PCa的诊断和治疗得到显著改善，从而代表着PCa管理的巨大进步。