CASIA OpenIR  > 毕业生  > 硕士学位论文
基于Monte Carlo 方法的生物自发光前向并行仿真平台
Alternative TitleA Parallel Code of Monte Carlo Simulation of Photon Transport in Bioluminescence Imaging
杨薇
Subtype工学硕士
Thesis Advisor杨鑫
2008-06-04
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline模式识别与智能系统
Keyword生物自发光成像 Monte Carlo 方法 并行 随机数发生器 主从通信 多光谱仿真 Bioluminescence Imaging Monte Carlo Method Parallel Random Number Generator Master/slave Communication Multi-spectral Simulation
AbstractMonte Carlo (MC) 方法是模拟生物组织内光子传输的最常用手段之一,其模拟结果的准确性也得到了多方面验证。实验室基于MC 方法,开发了生物自发光前向仿真平台MOSE(Molecular Optical Simulation Environment),可以支持二维、三维解析模型和真实环境下的任意形状面片模型中的光子传输仿真。 由于样本之间的独立性,MC 模拟有非常优良的并行品质。MC 方法有很多优点,而一个最大的缺点就是计算量大。并行计算恰好可以弥补这一缺点,使很多单机不能承受的模拟可以通过并行化得以快速地解决。本文就是将并行计 算和MC 算法结合起来,提高仿真的速度。 具体来说,本文的工作主要集中在两个方面: 1) 研发一个并行仿真平台ParaMose,即将并行机制引入到Console 版的生物自发光前向仿真程序中形成并行仿真平台ParaMose。并行机制包括,并行环境的选取、并行随机数发生器、动态任务分配、主从通信等。另外还包括多光谱的应用,并行仿真结果的处理等。 2) 针对并行仿真平台ParaMose 进行一系列验证实验和应用。首先对并行仿真的结果进行定量的验证,然后还对ParaMose 的并行性能进行研究,分别与串行软件MOSE 和国际上的知名软件triMC3D 进行对比,最后简单介绍了ParaMose 进行非匀质并行仿真、多光谱仿真、大光子数仿 真方面的应用。
Other AbstractMonte Carlo (MC) has been widely used to simulate light transport in tissues for various applications and gone through several improvements. Based on the Monte Carlo method, a platform named MOSE (Molecular Optical Simulation Environment) was developed in our lab, which could support the simulation of photon transport in 2D/3D geometries or arbitrary-shaped phantoms described with triangle meshes. Monte Carlo simulation has very good parallel quality because of its statistical independence. MC has many advantages, but one biggest disadvantage is that direct Monte Carlo simulation demands large computational capacity. However, parallel computing can make up this point. Many simulations of large capacity that the single computer can not bear can be executed gracefully on PC clusters through parallel computing. The main work of this dissertation is to combine parallel computing with Monte Carlo method to elevate simulation speed. Specifically, the work of this dissertation focuses on two aspects. First, a parallel code of Monte Carlo simulation for the photon transport in bioluminescence imaging is developed, i.e., the parallel mechanism is introduced into the console Monte Carlo code forming a parallel simulation platform named ParaMose. The parallel mechanism includes the selection of appropriate parallel environment, parallel random number generator, dynamic task allocation, master/slave communication, etc. Besides, there is also the application of multi-spectral photon transport simulation, combination of the parallel simulation results. Second, a series of experiments of validation and application on ParaMose are conducted. The parallel simulation results of ParaMose are validated quantitatively, and then the parallel performance of ParaMose is studied. ParaMose is also compared with serial Monte Carlo simulation via MOSE and parallel simulation via software triMC3D. Finally, the applications of ParaMose to a heterogeneous phantom, multi-spectral simulation and large numbers of photon simulation are presented.
shelfnumXWLW1263
Other Identifier200528014628056
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/7462
Collection毕业生_硕士学位论文
Recommended Citation
GB/T 7714
杨薇. 基于Monte Carlo 方法的生物自发光前向并行仿真平台[D]. 中国科学院自动化研究所. 中国科学院研究生院,2008.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20052801462805(1878KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[杨薇]'s Articles
Baidu academic
Similar articles in Baidu academic
[杨薇]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[杨薇]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.