Virtual dose mapping for radiation sterilization

项目摘要 这笔赠款是对PA-22-176的响应，PA-22-176是NIH、FDA和CDC的综合呼吁。FDA在其中表示 开发有助于监管决策的计算机建模工具的必要性。 灭菌是医疗器械开发的最后一步，灭菌验证是一项规范性的工作。 安宁。我们正在开发一种蒙特卡罗建模工具，它可以生成辐射的虚拟剂量图 消毒，从而帮助医疗公司做出关于消毒的重要决定，这也将影响 在医疗器械开发周期早期的调控途径。因为建模只需要一个CAD模型 而不是制造的设备(合同消毒器的剂量测绘需要)，mod- Elling可以在早期使用，以迭代设备及其包装的设计。这样的建模工具将允许 医疗器械公司缩短并降低灭菌方面的设计过程风险。 像Geant4这样的蒙特卡罗模型库能够精确地模拟辐射传输和剂量 证词。正因为如此，Geant4被广泛应用于高能物理和辐射治疗，但到目前为止，它 很少用于辐射灭菌。原因是缺乏用户友好性，这源于它的 灵活的体系结构和较慢的模拟速度。此外，Geant4不提供加载 完整的CAD模型，因此对整个医疗设备建模相当繁琐。 在FDA SBIR拨款的第一阶段，我们展示了Geant4可以部署在云中。我们已经开发出一种 用于配置模拟和查看建模结果的Web用户界面原型。我们的云工具可以加载 完整的CAD模型和精确复制测量的剂量图。我们展示了100倍的模拟增长 由于云中的无缝扩展，速度比单台计算机更快。 在第二阶段，我们将在第一阶段取得的成就的基础上，将该工具扩展到能够模拟剂量 X射线和伽马灭菌的地图。我们将开发可以计算的参数辐射源模型。 使用合同灭菌器提供的合格数据进行标记。我们将优化我们的用户界面，并提供 用户能够比较不同的辐射灭菌技术。我们还将实施以下能力 模型虚拟剂量计，便于与测量结果进行直接比较。在用户研究中，我们将展示 该工具用户友好，可用于简化监管备案所需的决策。 我们的建模工具将面向医疗设备公司的工程师，不需要 软件工程或辐射物理方面的高级背景。相反，工程师将被引导通过 模拟过程中的简单步骤。工程师可以在用户友好的查看器中查看结果，并 确定剂量均匀率。我们预计，我们的工具将为医疗设备提供显著的好处 开发社区。

Project Summary This grant is in response to PA-22-176, an omnibus call from NIH, FDA, and CDC. Therein the FDA expresses the need to develop computer modelling tools that help with regulatory decision making. Sterilization is the terminal step of medical device development, and sterilization validation is a regulatory re- quirement. We are developing a Monte Carlo modelling tool that can produce virtual dose maps for radiation sterilization, thereby helping medical companies make important decisions about sterilization that will also affect the regulatory pathway early in the medical device development cycle. As modelling only requires a CAD model of the device and not a manufactured device (which is required for dose mapping at a contract sterilizer), mod- elling can be used early to iterate on the design of the device and its packing. Such a modelling tool will allow medical device companies to shorten and de-risk the design process with respect to sterilization. Monte Carlo modelling libraries such as Geant4 have the ability to accurately model radiation transport and dose deposition. Because of this, Geant4 is widely used in high energy physics and radiation treatment, but so far it is rarely used for radiation sterilization. The reason for this is the lack of user friendliness, which stems from its flexible architecture and the slow simulation speeds. Furthermore, Geant4 does not provide the ability to load full CAD models, and thus makes it rather cumbersome to model entire medical devices. In Phase I of this FDA SBIR grant, we showed that Geant4 can be deployed in the cloud. We have developed a prototype of a web user interface to configure simulations and to review modeled results. Our cloud tool can load full CAD models and accurately reproduce measured dose maps. We showed a 100x increase of simulation speed compared to a single computer due to seamless scaling in the cloud. In Phase II we will build on our accomplishments of Phase I and extend this tool to be capable of modeling dose maps from X-ray and gamma sterilization. We will develop parametric radiation source models that can be cali- brated with qualification data provided by the contract sterilizer. We will refine our user interface and give the user the ability to compare the different radiation sterilization technologies. We will also implement the ability to model virtual dosimeters to facilitate direct comparison with measurements. In user studies we will show that the tool is user-friendly and can be used to streamline decisions necessary for regulatory filings. Our modelling tool will be geared towards engineers at medical device companies and does not require an advanced background in software engineering or radiation physics. Rather, the engineer will be guided through simple steps through the simulation process. The engineer can review the results in a user-friendly viewer and determine the dose uniformity ratio. We anticipate that our tool will offer significant benefits to the medical device development community.