低场核磁共振信号处理方法研究-资料分享下载-无忧分享网

中图分类号：

TP391.1

低场核磁共振信号处理方法研究

生物医学工程

李新军

聂生东教授

二○一一年十二月

学校代码：10252

学号：092461147

上海理工大学硕士学位论文

低场核磁共振信号处理方法研究

姓名李新军

系别医疗器械与食品学院

专业生物医学工程

研究方向医学图像处理

指导教师聂生东教授

学位论文完成日期 2011年12 月

University of Shanghai for Science and Technology

Master Dissertation

Signal Processing of Low-field Nuclear

Magnetic Resonance

Name Li Xinjun

Department College of Medical Instrument and Food

Engineering

Specialty Biomedical Engineering

Research Direction Medical Image Processing

Supervisor Professor Nie Shengdong

Complete Data December 2011

学位论文版权使用授权书

本学位论文作者完全了解学校有关保留、使用学位论文的规定，

同意学位论文保留并向国家有关部门或机构送交论文的复印件和电子

版。允许论文被查阅和借阅。本人授权上海理工大学可以将本学位论

文的全部内容或部分内容编入有关数据库进行检索，可以采用影印、

缩印或扫描等复制手段保存和汇编本学位论文。

本学位论文属于

学位论文作者签名：指导教师签名：

年月日年月日

保密年 □

不保密 □

摘要

随着核磁共振技术的发展，低场核磁共振已经能够用于检测样品中组分的种

类及含量等，具有测量方便、快速、无损、无毒副作用等优点，并且制造、维护、

应用成本低，在生物能源、纺织、环保、粮油食品、农业与生物、石油勘探、石

油化工、聚合物材料等方面有很广阔的应用前景。为了满足这些领域的需求，必

须对采集到的核磁共振信号进行反演等处理以提取信号中的有用信息。

一维反演技术在石油测井领域开展研究的时间最早，成果最多，而在其他领

域的研究仍未完全展开。经过一维反演可得到用于分析样品的组成及性质的核磁

共振一维谱，当其中多种组分的谱线相互叠加时，会使检测结果出现很大误差，

二维反演技术可很好的弥补这一缺陷。

近几年国外具备二维反演功能的低场核磁共振仪器已经商业化，而国内这方

面的研究相对落后，仍然以一维反演为主，二维反演尚处于起步阶段，而且由于

技术封锁、商业机密等原因，国内外公开的相关研究资料很少。本文在分析了石

油测井领域一维反演技术的基础上，结合国内外公开的有限的二维反演资料，提

出了一种新的扩散系数—横向弛豫时间(D—T2)二维反演算法。

本文首先对目前公开的一维反演算法进行详尽的分析，然后结合用于一维反

演的改进的奇异值分解算法和联合迭代修正算法的思想，对二维反演方程中的核

心函数矩阵进行分解，最后经过大量实验和理论研究确定反演过程中各个参数的

设置，并根据分解的矩阵进行 D—T2二维反演。

大量实验证明，本文算法能够得到理想的结果，并且所做的改进能够大幅度

降低计算量，减少消耗的计算机内存，提高 D—T2二维反演的速度。

关键词：核磁共振扩散系数横向弛豫时间二维反演

ABSTRACT

With the development of Nuclear Magnetic Resonance (NMR), Low-field NMR

has been able to detect the types and contents of sample compositions, and it is

convenient, fast and non-destructive without side effects. As the NMR equipments cost

little money to manufacturing, maintance and using, it has a very broad application

prospect in bio-energy, textile, environmental protection, foodstuff, agriculture, oil

exploration, petrochemicals, and polymer materials and so on. To meet all needs of

these areas, the signal collected with NMR equipment must be processed with inversion

to extract useful information.

The earliest research of one-dimension inversion is in Nuclear Magnetic

Resonance oil logging with most research results while the research in other areas has

not yet been expanded. One-dimension NMR spectra got from one-dimension inversion

can be used to analyze the composition and characteristic of samples. The conclusion

got from the one-dimension spectra should be wrong if there are some spectral lines of

ingredients in samples are superimposed, and the two-dimension spectra can solve the

problem effectively.

Some low-field NMR equipment with two-dimension inversion function

manufactured by foreign companies have been sold in market, but the domestic

technology of two-dimension inversion is backward, most researches in this area are

mainly the one-dimension inversion, and two-dimension inversion is still in its infancy.

With the reasons of technology blockade, trade secrets protection and so on, there are

only a little domestic and foreign public technical information can be got on the internet.

In this paper, we introduce a new diffusion-T2 relaxation two-dimension inversion

algorithm, based on the analysis of one-dimension inversion in Nuclear Magnetic

Resonance oil logging and the limited domestic and foreign public technical

information.

We make an elaborate analysis of the public one-dimension inversion algorithm

first. Then combined of the idea of improved Singular Value Decomposition algorithm

and the joint iterative correction algorithm, we divide the core function matrix into two

small matrixes. At last, doing a large number of experiments and theory researches we

get some parameters’ value. Last we introduce a new diffusion-T2 relaxation

two-dimension inversion algorithm.

A large number of experiments show that the improvement of the two-dimension

inversion algorithm introduced in this paper can reduce the compute resource, increase

the speed of the inversion, and using this algorithm, we can get more desired results.

Key Words：NMR, Diffusion, T2 relaxation, 2D Inversion

摘要

ABSTRACT

第一章绪论 ......................................................... 1

§1.1 研究背景 ..................................................... 1

§1.2 问题描述 ..................................................... 2

§1.3 主要研究内容 ................................................. 3

§1.4 本文结构 ..................................................... 4

第二章低场核磁共振简介 ............................................. 5

§2.1 核磁共振物理基础 ............................................. 5

§2.1.1 核磁共振物质基础 ........................................ 5

§2.1.2 拉莫尔进动 (Larmor Precession) .......................... 5

§2.1.3 横向弛豫时间 T2 .......................................... 6

§2.2 自由感应衰减信号（FID） ...................................... 7

§2.2.1 旋转坐标系 .............................................. 7

§2.2.2 自由感应衰减 ............................................ 7

§2.3 自旋回波 ..................................................... 8

§2.4 CPMG 脉冲序列 ................................................ 9

§2.5 扩散系数 D .................................................. 10

§2.5.1 扩散现象 ............................................... 10

§2.5.2 NMR 中宏观扩散运动的产生 ............................... 10

§2.5.3 扩散对 NMR 信号的影响 ................................... 11

第三章低场核磁共振数据反演 ........................................ 12

§3.1 反演简介及应用 .............................................. 12

§3.2 一维反演 .................................................... 12

§3.2.1 一维反演简介 ............................................ 12

§3.2.2 横向弛豫 T2布点 .......................................... 15

§3.2.3 CPMG 序列信号信噪比 SNR 的计算 ........................... 15

§3.2.4 常用的一维反演方法 ...................................... 16

§3.3 扩散-横向弛豫反演(D-T2反演) ................................. 25

§3.3.1 引言 ................................................... 25

§3.3.2 国内外进展 ............................................. 27

§3.3.3 二维反演简介 ........................................... 27

§3.3.4 现有 D-T2反演数据采集方式 ............................... 28

§3.3.5 扩散系数 D布点方式 ..................................... 31

§3.3.6 基线计算 ............................................... 32

§3.3.7 现有 D—T2反演方法 ....................................... 33

§3.3.8 本文采用的方法 .......................................... 34

§3.3.9 算法分析 ................................................ 37

§3.3.10 存在的问题 ............................................. 37

第四章结果与讨论 .................................................. 39

低场核磁共振信号处理方法研究

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

推荐作者

热门标签

举报选择: