Dual kV/MV Imaging for Metal Artifact Reduction

描述（由申请人提供）：现代癌症放疗在很大程度上依赖于计算机化的逆治疗计划，该计划通常使用诊断CT图像作为输入数据来描绘解剖结构并计算优化的调强放疗（IMRT）计划。图像引导放射治疗也需要良好的CT图像质量来验证患者的定位和识别肿瘤的缩小。然而，金属的存在，例如患者的髋关节植入物或牙科手术，会导致诊断和锥束kV-CT的图像伪影，严重影响治疗计划。由于金属对高能光子的吸收相对较低，特高压CT在很大程度上没有这种金属伪影，但代价是软组织描绘差，这对准确的治疗计划至关重要。我们的假设是，将诊断锥束CT成像与有限的巨电压CT数据采集相结合，可以以成本效益和剂量效率高的方式，在很大程度上消除复合CT数据集中的这些金属伪影。我们将设计并优化一种新型的、高DQE的、薄条的MV闪烁体，并将其放置在标准的电子传送门成像设备上，结合目标多叶准直器控制。这一假设的显著改善，图像引导放射治疗干预的多模式途径将测试以下具体目的：1)开发DQE(0) > 25%，分辨率为8 lp/cm的MV条形检测器；2)开发用于硬件控制、图像校正和重建的新软件，在小于100 mGy的联合剂量下提供120 Hounsfield单位精度；3)在至少80例癌症患者的初步体内临床研究中验证了放射治疗方案的改进。这些进展迅速转化为临床实践，将大大提高IMRT计划和图像指导的准确性，对于那些在其治疗区域内或附近有金属物体的患者，将显著降低发病率并改善局部肿瘤控制。这项研究的潜在影响超出了这里列出的应用范围，包括受金属植入物（如脊柱棒）影响的其他治疗领域，以及骨科植入物附近的定量CT成像。

DESCRIPTION (provided by applicant): Modern radiotherapy of cancer relies heavily on computerized inverse treatment planning, which typically uses diagnostic CT images as input data to delineate anatomy and compute optimized intensity modulated radiotherapy (IMRT) plans. Image guided radiotherapy also requires good CT image quality to verify patient positioning and to identify tumor shrinkage. However the presence of metal, for example hip implants or dental work in the patient, causes image artifacts in diagnostic and conebeam kV-CT that severely impact treatment planning. Megavoltage CT is largely free of such metal artifacts due to the relatively low absorption of high energy photons by metal, but at the expense of poor soft tissue delineation essential to accurate treatment planning. Our hypothesis is that the combination of diagnostic cone beam CT imaging with limited megavoltage CT data acquisition can largely eliminate these metal artifacts from a composite CT data set, in a cost effective and dose-efficient manner. We will design and optimize a novel, high DQE, thin strip, MV scintillator placed on a standard Electronic Portal Imaging Device in combination with targeted multi-leaf collimator control. This hypothesis of significantly improved, image guided radiotherapy intervention in a multi-modal approach will be tested with the following specific aims: 1) develop the MV strip detector with a DQE(0) > 25% and a resolving power of 8 lp/cm, 2) develop new software for hardware control, image correction and reconstruction that provides 120 Hounsfield unit accuracy from a combined dose of less than 100 mGy, and 3) validate the radiation therapy treatment plan improvement in a preliminary in-vivo clinical study on at least 80 cancer patients. The rapid translation of these advances into clinical practice should greatly improve the accuracy of IMRT planning and image guidance, for those patients with metal object in or near their treatment fields, leading to significantly reduced morbidity combined with improved local tumor control. The potential impact of this research extends beyond the applications listed here, to include other therapy treatment fields affected by metallic implants such as spinal rods, as well as quantitative CT imaging in the vicinity of orthopaedic implants. PUBLIC HEALTH RELEVANCE: Modern radiation therapy treatment of cancer depends on the automated computer generation of complex radiation delivery treatment plans to ensure that the correct dose is delivered only to the tumor. The accuracy of these plans is critically dependent on the clarity of the diagnostic X-ray CT images of the cancer region. A large fraction of the patients who receive radiation therapy of head and neck cancer have teeth filled with metal amalgams or covered with gold crowns, and those with prostate cancer may have artificial hips, both of which degrade the quality of the CT images and the accuracy of the resulting treatment plans. We plan to overcome this limitation by designing and building a new, combined kV/MV imaging system that provides excellent images without requiring a significant increase in dose to the patient.