Quantitative breast cancer risk index from routine 3-D imaging

描述：此应用程序响应PA-10-026，“癌症风险和预后预测模型的开发、应用和评估(R21)”。这项研究针对疾病和器官，旨在开发和验证一种准确的方法来确定乳房密度，乳房密度是乳腺癌的已知风险因素。开发一种准确的估计乳房密度的方法对于确保癌症风险和预后预测模型的准确性至关重要。几项研究表明，乳房密度与乳腺癌风险之间存在关联。荟萃分析发现，纤维腺含量大于75%的女性与纤维腺含量低于5%的女性相比，其相对危险度约为4.7。就相对风险而言，乳房密度现在被认为是仅次于年龄和BRCA突变的第三高风险因素。认识到乳房密度作为风险因素的重要性，以及乳房致密的女性对乳房X光检查的敏感度降低，几个州现在规定，接受筛查乳房X光检查的女性必须被告知她的乳房密度。在美国，乳房密度是由口译放射科医生根据美国放射学、乳房成像报告和数据系统报告的四个类别进行报告的。已经表明，放射科医生之间对这种分类分配的一致意见只是中等的。因此，有必要开发准确的定量技术来估计乳房密度。虽然一些研究使用乳房投影面积的乳房密度定量估计，通常被称为乳房X光检查密度的百分比，但最近的研究表明，乳房密度的体积估计更准确地预测乳腺癌风险。在这项研究中，我们建议开发和评估一种定量算法，作为一种工具，以数字乳房断层合成(DBT)提供的3-D图像为基础，基于实质纹理分析来估计乳房体积密度和相关测量。至少有一家DBT系统制造商已经获得FDA批准用于常规临床，几家制造商正在努力争取FDA的批准。因此，开发的定量工具将被设计成与多家供应商的DBT系统兼容，促进其在临床研究中的广泛使用。这对于从可能使用不同供应商的DBT系统的多个站点汇集图像数据的研究尤其重要。这项研究的具体目的包括开发定量工具，评估其定量准确性，并进行可行性研究，目的是在以乳腺MRI为真理的临床人群中验证其准确性。经过验证的基于3D成像的定量工具可用于未来的研究，以准确确定与乳房密度相关的癌症风险，评估药物干预对乳房密度和癌症风险的影响，并开发个性化的乳腺癌筛查方案。

DESCRIPTION: This application is responsive to PA-10-026, "Development, Application, and Evaluation of Prediction Models for Cancer Risk and Prognosis (R21)". This research is disease and organ-specific and is aimed at developing and validating an accurate method for determining breast density, a known risk factor for breast cancer. Developing an accurate method for estimating breast density is essential to ensure the accuracy of prediction models for cancer risk and prognosis. Several studies have shown the association between breast density and breast cancer risk. A meta-analysis observed that the relative risk was ~4.7 for women with greater than 75% fibroglandular content compared to those with less than 5% fibroglandular content. Breast density is now considered the third highest risk factor in terms of relative risk after age and BRCA mutation. Recognizing the significance of breast density as a risk factor and the reduced sensitivity of mammography for women with dense breasts, several states now regulate that a woman undergoing screening mammography is informed of her breast density. In the United States, breast density is reported as per the American College of Radiology, Breast Imaging Reporting and Data System by the interpreting Radiologist that uses four categories. It is has been shown that agreement between Radiologists for such categorical assignment is only moderate. Hence, there is a need to develop accurate quantitative techniques for estimating breast density. While some studies use quantitative estimates of breast density from the projected area of the breast, often referred to as percent mammographic density, recent research has shown that volumetric estimates of breast density are more accurate predictors of breast cancer risk. In this research, we propose to develop and evaluate a quantitative algorithm that serves as a tool for the estimation of volumetric breast density and associated measures based on parenchymal texture analysis using 3-D images provided by digital breast tomosynthesis (DBT). At least one DBT system manufacturer has obtained FDA-approval for routine clinical use and several manufacturers are working towards FDA approval. Hence, the developed quantitative tool will be designed to be compatible with DBT systems from multiple vendors, facilitating its widespread use for clinical studies. This is particularly important for studies where image data are pooled from multiple sites that may use DBT systems from different vendors. The specific aims of this research include developing the quantitative tool, evaluating its quantitative accuracy, and conducting a feasibility study aimed a demonstrating its accuracy in a clinical population with breast MRI serving as the truth. The validated 3-D imaging based quantitative tool can be used in future studies to accurately determine the cancer risk associated with breast density, to assess the effect of pharmacologic intervention on breast density and cancer risk, and to develop personalized breast cancer screening regimens.