Administrative Supplement: Development of functional magnetic resonance imaging-guided adaptive radiotherapy for head and neck cancer patients using novel MR-Linac device

项目摘要/摘要：为癌症提供剂量，同时保留正常组织是最终目标 在放射治疗(RT)中，尤其是在头颈部。通过确定哪些肿瘤更有可能 在治疗的早期反应，以及耐药肿瘤的亚体积，RT计划可以每天改变以 利用肿瘤中的生物改变，以同等概率减少副作用 治愈的方法。功能成像技术已在临床系列中显示出在早期鉴别方面的实用价值 头颈癌(HNC)放射治疗的应答者，以及识别辐射抗性 疾病治疗后。这些功能成像技术可以用来主动适应放射治疗 在放射治疗过程中出现的频率很高。 与我们的工业合作伙伴(Elekta医疗系统)合作，我们的集团最近获得了 NIH R01拨款(5R01DE028290-03，FREEDOMM-RT)，用于开发硬件、软件、技术和 HNC患者功能影像引导放射治疗的质量保证基础设施。由此产生的高频 来自该项目的解剖和功能成像数据，以及其他MRI数据源 形成了一个史无前例的大数据语料库，用于MRI引导的自适应RT。因此， 在本补充资料中，我们建议对这些数据进行有选择的整理、注释和传播，以便于 社区驱动的人工智能(AI)模型构建工作，以便更容易地翻译MR引导的 将放射治疗技术应用于临床。 拟议的为期一年的补编由数据整理和数据挑战执行工作组成。 具体地说，我们将在多个时间点和 生成感兴趣的相应分割区域；剂量学、人口学和临床数据 为每一位病人策划。这些基准数据集将被匿名并传输到癌症 影像归档以供公众重复使用，从而促进研究社区开发强大的以RT为中心的人工智能 项目。此外，为了促进社区参与我们的新基准数据集，我们将发起一个 一连串的AI“数据挑战”。通过这些挑战，我们将直接培育新的AI创新来解决 临床上相关的RT问题。 该项目的成功完成将使生物医学数据现代化和集成化 为公众使用RT数据构建人工智能模型的生态系统。此外，拟议的基准数据集将 为使用人工智能实现HNC患者个性化医疗的长期目标提供基础 减少口腔牙科后遗症，同时保持出色的治愈率，直接补充 家长助学金。最后，本补充剂将通过使患者能够描述 改善治疗干预和人工智能技术的下游翻译应用的恶性肿瘤。

PROJECT SUMMARY/ABSTRACT: Delivering dose to cancers while sparing normal tissue is the ultimate goal in radiotherapy (RT) treatment, especially in the head and neck. By identifying tumors which are more likely to respond early in treatment, as well as subvolumes of resistant tumor, RT plans could be changed each day to take advantage of biological alteration in the tumor, resulting in reduced side effects with equivalent probability of cure. Functional imaging techniques have demonstrated utility in clinical series in discriminating early responders to radiation therapy in head and neck cancer (HNC), as well as identifying radiation resistant disease post-therapy. These functional imaging techniques could be utilized to actively adapt radiation therapy with high frequency during the radiation treatment course. In tandem with our industrial partner (Elekta Medical Systems), our group has recently been awarded an NIH R01 grant (5R01DE028290-03, FREEDOMM-RT) to develop the hardware, software, technical, and quality assurance infrastructure for functional image-guided RT for HNC patients. The resulting high-frequency anatomical and functional imaging data derived from this project, in addition to additional MRI data sources from our institution forms a corpus of unprecedented novel “big data” for MRI-guided adaptive RT. Therefore, in this supplement, we propose the selective curation, annotation, and dissemination of these data to facilitate community-driven artificial intelligence (AI) model building efforts in order to more readily translate MR-guided RT technologies into the clinic. The proposed one-year supplement is composed of data curation and data challenge execution efforts. Specifically, we will curate high-quality anatomical and functional MRI sequences at multiple timepoints and generate corresponding segmentations regions of interest; dosimetric, demographic, and clinical data will be curated for each patient. These benchmark datasets will be anonymized and transmitted to The Cancer Imaging Archive for public re-use, thereby fostering the research community to develop robust RT-centric AI projects. Additionally, to facilitate community engagement with our novel benchmark datasets, we will initiate a series of AI “data challenges”. Through these challenges we will directly foster novel AI innovation to solve clinically relevant RT problems. Successful completion of this project will enable a modernized and integrated biomedical data ecosystem for public use of RT data for AI model building. Moreover, the proposed benchmark datasets will provide a foundation to achieve the long-term goal of personalized medicine for HNC patients using AI to reduce oro-dental sequelae while maintaining excellent cure rates, directly complementing the goals of the parent grant. Finally, this supplement will positively impact patients by enabling the characterization of malignancy for improved therapeutic intervention and downstream translational application of AI technologies.