Low dose cone beam CT for image guided adaptive radiotherapy

描述（申请人提供）：锥形束计算机断层扫描（CBCT）已广泛应用于图像引导放射治疗（IGRT）和适应性放射治疗（ART），以获取更新的患者几何形状，以精确靶向和治疗适应。然而，在治疗过程中反复使用CBCT引起了对给予患者的过量x射线成像剂量的严重关注，这极大地限制了现代放疗潜力的最大利用。特别是对于儿科患者，这种担忧已经禁止使用IGRT和ART，导致治疗结果受损。基于压缩感知技术的先进迭代重建算法在从少量和/或噪声投影中重建CBCT图像方面显示出巨大的能力，从而大大降低了成像剂量。然而，这些算法的计算效率非常低，因此不能用于大多数临床应用。我们最近在开发一种创新的CBCT重建算法方面取得了突破，该算法具有完美的数学结构，可以在图形处理单元（GPU）平台上并行化。我们的初步结果表明，与现有的迭代算法相比，我们的效率可以提高100倍，与目前的临床标准相比，我们的成像剂量可以降低40~100倍。我们的目标是通过追求以下两个具体目标，将这一有前途的算法发展成为临床功能的CBCT重建系统，该系统可以在极低的辐射剂量（<当前剂量的1%）和高速（< 5秒）下产生高质量的CBCT图像：我们将开发一种基于gpu的系统，以超低辐射剂量和超高速重建高质量的CBCT图像。SA2。我们将通过一系列数值、模拟和患者研究来评估该系统，展示成像剂量降低的增益，并在各种临床条件下建立临床方案。拟建项目完成后，将系统开发并评估一套超低剂量、超快性能的临床可用CBCT重建系统。该系统的临床应用将使大量接受现代放射治疗的患者受益。特别是，我们的工作将首次使IGRT和ART临床应用于儿科患者。

DESCRIPTION (provided by applicant): Cone-beam computed tomography (CBCT) has been broadly used in image guided radiation therapy (IGRT) and adaptive radiation therapy (ART), to acquire the updated patient's geometry for precise targeting and treatment adaptation. However, the repeated use of CBCT during a treatment course has raised a serious concern on excessive x-ray imaging doses delivered to patients, which has greatly limited the maximal exploitation of the potential of modern radiotherapy. Especially for pediatric patients, this concern has prohibited the use of IGRT and ART, leading to compromised treatment outcome. Advanced iterative reconstruction algorithms, based on compressed sensing techniques, have demonstrated tremendous power in reconstructing CBCT images from very few and/or noisy projections, resulting in dramatically reduced imaging dose. However, these algorithms are very computationally inefficient and thus cannot be used in most clinical applications. We have recently made a breakthrough in developing an innovative CBCT reconstruction algorithm with a mathematical structure perfect for parallelization on a graphics processing unit (GPU) platform. Our preliminary results have shown that we can improve the efficiency by a factor of 100 over existing iterative algorithms and reduce the imaging dose by factor of 40~100 compared to the current clinical standard. Our goal is to develop this promising algorithm into a clinically functioning CBCT reconstruction system which can produce high quality CBCT images at extremely low radiation dose (<1% of the current dose) and high speed (< 5 seconds), by pursuing the following two specific aims: SA1. We will develop a GPU-based system to reconstruct high quality CBCT images at ultra-low radiation dose and ultra-high speed. SA2. We will evaluate the system through a series of numerical, phantom, and patient studies, demonstrate the gain in imaging dose reduction, and establish clinical protocols under various clinical conditions. Upon the completion of the proposed project, a clinically ready-to-use CBCT reconstruction system with ultra-low dose and ultra-fast performance will have been systematically developed and evaluated. Clinical introduction of such a system will significantly benefit a large number of patients receiving modern radiotherapy. Especially, our work will for the first time make IGRT and ART clinically available for pediatric patients.