Constrained Disentanglement (CODE) Network for CT Metal Artifact Reduction in Radiation Therapy

用于减少放射治疗中 CT 金属伪影的约束解缠结 (CODE) 网络

• 批准号: 10184493
• 负责人: Bruno De Man
• 金额: $ 259.38万
• 依托单位: RENSSELAER POLYTECHNIC INSTITUTE
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-09-15 至 2024-09-14
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10184493
• 关键词: Affect; Aging; Artificial Intelligence; Benchmarking; Cancer Patient; Cause of Death; Cessation of life; Clinic; Clinical; Communities; Computer software; Data; Dimensions; Estimation Techniques; Evaluation; General Hospitals; Goals; Human; Image; Implant; Institutes; Lead; Learning; Malignant Neoplasms; Massachusetts; Metals; Methods; Modeling; Modification; Morphologic artifacts; Network-based; Output; Patient-Focused Outcomes; Patients; Performance; Photons; Positioning Attribute; Protons; Radiation therapy; Recording of previous events; Records; Reporting; Research; Residual state; Supervision; Techniques; Technology; Technology Transfer; Testing; Therapeutic Studies; Time; Trace metal; Translating; Translations; Treatment outcome; World Health Organization; X-Ray Computed Tomography; base; clinical application; clinical imaging; clinical translation; computerized data processing; deep learning; dosimetry; experimental study; image reconstruction; imaging Segmentation; imaging modality; imaging software; improved; industry partner; innovation; learning network; learning strategy; man; meetings; open source; proton therapy; reconstruction; tomography; treatment planning

 ABSTRACT The World Health Organization reported that cancer is the second leading cause of death globally and is re- sponsible for 9.6 million deaths in 2018. Approximately 50% of all cancer patients receive radiation therapy (RT). Many of them have metal implants, which induce image artifacts in the treatment planning CT images and compromise or preclude treatment in an estimated 15% of all radiation therapy patients. Despite extensive CT metal artifact reduction (MAR) research it remains one of the long-standing challenges in the CT field, without a clinically satisfactory solution. The overall goal of this project is to develop cutting-edge deep learning imaging methods and software solutions for commercial CT scanners to eliminate CT metal artifacts in general and improve RT in particular. We propose a three-pronged approach to systematically tackle this challenge in three specific aims: (1) adversarial learning techniques for estimation of sinogram missing data and metal traces; (2) constrained disentanglement (CODE) networks to remove CT image artifacts during image reconstruction, through post-processing, and in both data and image domains; and (3) systematic evaluation of our proposed CT MAR techniques and clinical translation into robust RT planning methods to maximize the RT treatment planning accuracy and thus improve patient outcomes. Our synergistic track records in CT MAR research, especially with deep imaging methods over the past three years, promises an unprecedented opportunity for a brand-new solution to CT MAR. For the first time we will integrate contemporary AI innovations in data preprocessing, image reconstruction, post-processing, observer studies and treatment planning synergistically in a unified data-driven framework, positioning this project uniquely to eliminate metal artifacts and their complications in radiation therapy. This project will be pursued through the long-term academic-industrial partnership among Dr. Ge Wang at Ren- sselaer Polytechnic Institute (RPI), Dr. Bruno De Man at GE Research Center (GRC), and Dr. Harald Paganetti at Massachusetts General Hospital (MGH). While our teams will collaborate closely through the whole project, GRC has a history of CT research and translation, including direct raw data processing, and will focus on Aim 1. RPI is a pioneering group in tomographic reconstruction, especially deep-learning-based CT imaging, and will lead Aim 2. The MGH team is at the forefront of radiation therapy research and will be responsible for Aim 3. Upon completion of this project, we will have redefined the state of the art of CT MAR, largely eliminating CT metal artifacts and substantially improving radiation therapy planning and delivery accuracy. With the above-proposed networks for CT MAR, metal artifacts will have been basically eliminated, targeting residual errors <10 HU for photon and proton therapy planning, with the goal of reducing the clinical diametric error to ±3% and the proton range error due to metal artifacts to <2mm. Since our approach is software-based and open-source, the path for technology transfer and clinical translation is clearly defined, as well tested before.

 摘要 世界卫生组织报告说，癌症是全球第二大死亡原因，并且正在重新成为 对2018年960万人的死亡负有责任。大约50%的癌症患者接受放射治疗 (RT)。其中许多都有金属植入物，这会在治疗计划的CT图像和 在所有放射治疗患者中，估计有15%的患者接受了妥协或排除治疗。尽管进行了大量的CT检查 金属伪影减少(MAR)研究它仍然是CT领域的长期挑战之一，如果没有 临床满意的解决方案。 该项目的总体目标是开发尖端的深度学习成像方法和软件解决方案 用于商业CT扫描仪，以消除一般的CT金属伪影，特别是改善RT。我们建议 系统地应对这一挑战的三管齐下的方法有三个具体目标：(1)对抗性学习 正弦图缺失数据和金属痕迹的估计技术；(2)约束解缠(代码) 用于在图像重建期间、通过后处理和在两个数据中去除CT图像伪影的网络 和图像域；以及(3)对我们建议的CT MAR技术和临床翻译的系统评估 引入稳健的RT计划方法，最大限度地提高RT治疗计划的准确性，从而提高患者的治疗质量 结果。我们在CT MAR研究方面的协同跟踪记录，特别是在 在过去的三年里，CT MAR迎来了一个前所未有的机遇，一个全新的解决方案。第一次 我们将整合当代人工智能在数据预处理、图像重建、后处理、 观察者研究和治疗规划在统一的数据驱动框架中协同，定位于 独一无二的项目，消除放射治疗中的金属伪影及其并发症。 该项目将通过任-王博士之间的长期学术-产业合作伙伴关系来进行。 Sselaer理工学院(RPI)、GE研究中心(GRC)的Bruno De Man博士和Harald Paganetti博士 在马萨诸塞州综合医院(MGH)。虽然我们的团队将在整个项目中密切合作， GRC有CT研究和翻译的历史，包括直接原始数据处理，并将专注于目标1。 RPI是断层重建，特别是基于深度学习的CT成像方面的先驱组织，并将 领导目标2。MGH团队处于放射治疗研究的前沿，将负责目标3。 在这个项目完成后，我们将重新定义CT MAR的技术状态，在很大程度上消除CT 金属伪影，并大大提高放射治疗计划和提供的准确性。与 以上建议的CT MAR网络，金属伪影将基本消除，目标是残留 光子和质子治疗计划的误差&lt；10HU，目标是将临床直径误差减少到 ±3%，金属伪影引起的质子射程误差<lt；2 mm。由于我们的方法是基于软件的， 开源，技术转移和临床翻译的路径被明确定义，之前也得到了很好的测试。

项目成果

Deep learning tomographic reconstruction through hierarchical decomposition of domain transforms.

• DOI: 10.1186/s42492-022-00127-y
• 发表时间: 2022-12-09
• 期刊: Visual computing for industry, biomedicine, and art

SPICE: Semantic Pseudo-Labeling for Image Clustering.

• DOI: 10.1109/tip.2022.3221290
• 发表时间: 2022
• 期刊: IEEE transactions on image processing : a publication of the IEEE Signal Processing Society

Development and tuning of models for accurate simulation of CT spatial resolution using CatSim.

使用 CatSim 开发和调整模型以精确模拟 CT 空间分辨率。

• DOI: 10.1088/1361-6560/ad2122
• 发表时间: 2024
• 期刊: Physics in medicine and biology
• 影响因子: 3.5
• 作者: Zhang,Jiayong;Wu,Mingye;FitzGerald,Paul;Araujo,Stephen;DeMan,Bruno
• 通讯作者: DeMan,Bruno

Deep Few-view High-resolution Photon-counting Extremity CT at Halved Dose for a Clinical Trial

• DOI: 10.48550/arxiv.2403.12331
• 发表时间: 2024-03
• 期刊: ArXiv
• 作者: Mengzhou Li;Chuang Niu;Ge Wang;Maya R. Amma;Krishna M. Chapagain;Stefan Gabrielson;Andrew Li;Kevin Jonker;Niels J. A. De Ruiter;Jennifer A. Clark;Phillip H. Butler;Anthony Butler;Hengyong Yu

Making "CASES" for AI in Medicine.

• DOI: 10.34133/bmef.0036
• 发表时间: 2024
• 期刊: BME frontiers
• 作者: Wang G