Improved optical Monte Carlo simulation through standardization, robustness, and training

通过标准化、鲁棒性和训练改进光学蒙特卡罗模拟

• 批准号: 10584410
• 负责人: Emilie Roncali
• 金额: $ 23.94万
• 依托单位: UNIVERSITY OF CALIFORNIA AT DAVIS
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-30 至 2023-09-29
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10584410
• 关键词: 3-Dimensional; Academia; Address; Administrative Supplement; Algorithms; Architecture; Award; Clinical Research; Code; Collaborations; Combined Modality Therapy; Communities; Computer software; Crystallization; Custom; Dependence; Development; Discipline of Nuclear Medicine; Documentation; Education; Feedback; Fostering; Future; Generations; Goals; Grant; Hybrids; Imaging technology; Industry; International; Manuals; Measures; Methods; Modeling; Monte Carlo Method; Multimodal Imaging; Optics; Parents; Peer Review; Philosophy; Photons; Physics; Positron-Emission Tomography; Process; Programming Languages; Publications; Pythons; Research; Scientist; Software Engineering; Software Tools; Source Code; Speed; Standardization; Structure; Surface; Time; Training; Translating; United States National Institutes of Health; Work; base; computerized tools; data exchange; design; detector; diagnostic tool; empowered; flexibility; imaging detector; improved; innovation; member; molecular imaging; new technology; nuclear imaging; open data; personalized medicine; portability; radiation detector; simulation; simulation software; software development; tool; user-friendly

Project Summary/ Abstract This application is an administrative supplement to grant R01 EB027130, which develops new optical Monte Carlo simulation tools for nuclear medicine in the opensource GATE/Geant4 simulators. The work proposed here will improve these tools through standardization, robustness, and training with the objective of enabling more users to fully utilize the tools we freely release. GATE/Geant4 is the main simulation platform in nuclear imaging and therapy. It includes optical transport in scintillators, but the optical models are highly inaccurate. We have developed and integrated into GATE a new optical model, the “LUT Davis model” that addresses this limitation. The two primary objectives of the parent award are 1) to develop and freely distribute computational tools to generate custom optical surface LUTs, and 2) to develop optical models for photon timing studies and establish a comprehensive simulation framework for detector timing optimization. We have developed a free standalone application for users to describe their custom surfaces, using custom code entirely developed by research scientists and trainees. This supplement will apply software engineering methods to improve the LUT Davis standalone app robustness, portability, documentation, and integration in GATE. Two specific aims are proposed that include software development and new training. Aim 1 will improve the robustness and flexibility of the LUT Davis app and enhance its responsiveness to the needs of the nuclear imaging research community. Refactoring the standalone app in one of the most utilized programming languages (Python) that is freely accessible will be more aligned with our open science philosophy and will offer the possibility to integrate the optical surface design into simulation frameworks developed by users. We anticipate this to increase the standalone usage, make the code more reusable, and further accelerate research. These software developments will be associated with new training dedicated to optical simulation. Aim 2 focuses on refactoring the API that calls the GATE LUT Davis model in the simulations. Our 2- pronged approach will restructure the current GATE source code workflow and prepare for the new GATE under development. Cleaner code will make simulations faster and more robust by reducing computation errors and streamlining processes. The new GATE will be entirely Python-based to speed up and ease data transfer from GATE to analysis tools, overall making it a faster, easier, and more agile simulation toolkit responsive to dynamic user needs and more compliant with software standards. Our objective is to increase the impact of these tools by enabling more users to fully utilize the functionalities we create and release. Empowering the scientific community with enhanced simulation software will accelerate the development of nuclear imaging technology.

项目总结/摘要 该申请是对授予R 01 EB 027130的行政补充，该申请开发了新的光学 在开源GATE/Geant 4模拟器中用于核医学的蒙特卡罗模拟工具。工作 这里提出的建议将通过标准化，健壮性和培训来改进这些工具， 我们的目标是让更多的用户能够充分利用我们免费发布的工具。 GATE/Geant 4是核成像和治疗领域的主要仿真平台。它包括光学传输 但是光学模型非常不准确。我们已经开发并集成到GATE中， 新的光学模型，“LUT戴维斯模型”，解决了这一限制。两个主要目标 父母奖是1）开发和免费分发计算工具，以产生定制的光学表面 LUT，以及2）开发用于光子定时研究的光学模型，并建立全面的模拟 用于检测器定时优化的框架。我们为用户开发了一个免费的独立应用程序， 使用完全由研究科学家和学员开发的自定义代码描述其自定义表面。 该补充将应用软件工程方法来改进LUT Davis独立应用程序 GATE中的健壮性、可移植性、文档和集成。 提出了两个具体目标，包括软件开发和新的培训。目标1将改善 LUT Davis应用程序的稳健性和灵活性，并增强其对核需求的响应能力 成像研究社区。在最常用的编程之一中重构独立应用程序 免费访问的语言（Python）将更加符合我们的开放科学理念， 提供了将光学表面设计集成到用户开发的仿真框架中的可能性。 我们预计这将增加独立使用，使代码更可重用，并进一步加速 research.这些软件开发将与专门用于光学模拟的新培训相关联。 目标2侧重于重构在模拟中调用GATE LUT Davis模型的API。我们的2- 一种多管齐下的方法将重组当前的GATE源代码工作流程，并为新的GATE做好准备 正在开发中。更简洁的代码将通过减少计算使模拟更快、更健壮 错误和简化流程。新的GATE将完全基于Python，以加速和简化数据 从GATE转移到分析工具，总体上使其成为一个更快、更容易和更灵活的仿真工具包 响应动态用户需求，更符合软件标准。 我们的目标是通过使更多用户能够充分利用这些功能来增加这些工具的影响 我们创造和释放。通过增强的模拟软件增强科学界的能力， 加速核成像技术的发展。