Multi-scale modeling of glioma for the prediction of treatment response, treatment monitoring and treatment allocation

项目摘要 计算多尺度建模是一个不断发展的研究领域，其目的是将整个幻灯片图像和 放射影像与相同患者的多组学分子图谱。多尺度模型显示 通过其预测临床结局（如预后）的能力以及通过预测可采取行动的 肿瘤的分子特性，例如EGFR的活性，EGFR是许多癌症中的主要药物靶标。电流 应用仅限于研究成像和分子数据之间的关联， 结果。从多尺度生物标志物中还不能获得可操作的信息。 我们建议开发一个多尺度建模框架，以支持治疗反应、治疗监测 脑肿瘤患者的治疗分配，重点是最具侵袭性的胶质瘤亚型，IDH 野生型高级别神经胶质瘤。在目标1中，我们将开发整合多尺度数据的信息学算法， 治疗反应。我们将利用我们在数据融合方面的专业知识，开发新的方法来整合多个 规模数据，以预测一线治疗反应。在目标2中，我们将开发允许组合的算法 诊断时的多尺度数据与治疗随访期间的多模态MR成像数据。我们将专注于 预测治疗反应和进展，以及我们是否可以在 放射科医生可以。在目标3中，我们将开发使用多尺度数据来预测药物靶点的算法 活性，并为对一线治疗产生耐药性的患者提供新药。我们将使用一个 公开可用的胶质瘤多尺度数据集的混合物，总计超过1000名患者，以及1600名 来自斯坦福大学医学中心的150名回顾性和前瞻性脑肿瘤患者。 将这些互补的数据源组合在多尺度数据融合框架中可以产生深远的影响。 通过揭示未知的协同作用和关系，为预测治疗结果做出贡献。更 具体而言，开发整合定量图像特征和分子数据的计算模型， 开发多尺度特征，通过研究， 生物标志物，准确预测治疗反应。容易，因为整个幻灯片图像和射线照相 成像是癌症患者常规诊断检查的一部分， 越来越多地用于临床工作流程，因此，如果可以找到可靠的多尺度签名， 治疗反应，转化为临床应用是可行的，包括优化临床招募 审判

Project summary Computational multi-scale modeling is a growing area of research that aims to link whole slide images and radiographic iamges with multi-omics molecular profiles of the same patients. Multi-scale modeling has shown its potential through its ability to predict clinical outcomes e.g. prognosis, and through predicting actionable molecular properties of tumors, e.g. the activity of EGFR, a major drug target in many cancers. Current applications are limited to study associations between imaging and molecular data, and predicting long term outcomes. No actionable information can be gained from multi-scale biomarkers yet. We propose to develop a multi-scale modeling framework to support treatment response, treatment monitoring and treatment allocation for patients with brain tumors, focusing on the most aggressive subtype of glioma, IDH wild-type high grade glioma. In Aim 1, we will develop informatics algorithms that integrate multi-scale data for treatment response. We will use our expertise in data fusion and develop novel approaches to integrate multi- scale data to predict first line treatment response. In Aim 2, we will develop algorithms that allow combining multi-scale data at diagnosis with multi-modal MR imaging data during treatment follow-up. We will focus on predicting treatment response and progression and whether we can predict these events earlier than radiologists can. In Aim 3, we will develop algorithms that use the multi-scale data to predict drug target activities and also suggest novel drugs for patients that become resistant to first line treatment. We will use a mixture of publicly available glioma multi-scale data sets totaling more than 1000 patients, and also 1600 retrospective and 150 prospective brain tumor patients from Stanford Medical Center. Combining these complementary data sources in a multi-scale framework for data fusion can have profound contributions toward predicting treatment outcomes by uncovering unknown synergies and relationships. More specifically, developing computational models integrating quantitative image features and molecular data to develop multi-scale signatures, holds the potential to translate in benefit to brain tumor patients by investigating biomarkers that accurately predict treatment response. Readily, because whole slide images and radiographic imaging is part of the routine diagnostic work-up of cancer patients and molecular data of brain tumors is increasingly being used in clinical workflows, therefore if reliable multi-scale signatures can be found reflecting treatment response, translation to clinical applications is feasible, including optimizing recruitment for clinical trials.