zoukankan      html  css  js  c++  java
  • 学习笔记171—疾病【离散变量】和年龄【连续变量】的ANOVA+NBS

    目的:分别探究疾病和年龄的主效应,以及它们的交互效应,并生成连通网络!

    环境:window7及以上, matlab2010及以上

    一、通过NBS软件实现;

    1)软件准备

        ① 软件下载连接:https://www.nitrc.org/frs/?group_id=629

        ② 软件安装:(参考软件操作手册: reference_manual_NBS_v1.2.pdf )

        

       ③ 软件界面:(在 matlab 命令窗口输入:NBS)

          

    2)数据准备:数据矩阵,设计矩阵,对比矩阵,结点三维坐标(画图使用),结点名称(画图使用,可选)

       ① 数据矩阵:

    CONNECTIVITY MATRICES:
    The connectivity matrices can be given either as a:
    1. Text file (*.txt), where a separate text file is used to store the connectivity matrix for each subject/observation (provide the path and name of only one of the text files and any others in the same directory will be identified automatically)
    2. Valid Matlab expression (type the expression)
    3. Matlab file (*.mat) containing a single variable
    The Matlab expression or the Matlab file must specify a numeric array with three dimensions, where concatenation across the observations/subjects is in the third dimension. Specifically, element (i,j,k) stores the connectivity value between node i and node j for observation/subject number k.  

    (Specify a symmetric connectivity matrix for each observation or subject. The connectivity matrices store the connectivity strength between each pair of network nodes. Optionally, specify a label for each node as well as its MNI coordinates expressed in millimetres.)

        ② 设计矩阵和对比矩阵:

            a. main effect (主效应,不回归年龄) :C1 和 C2 是疾病的主效应

             b. Main effect (主效应, 回归年龄): C1和C2 是疾病的主效应(回归年龄后),C3和C4是年龄的主效应

           c. Interaction effect (疾病和年龄的交互效应): C1 和C2是疾病和年龄的交互效应

      ③ 结点三维坐标:

    NODE COORDINATES (MNI) [optional]:
    Node coordinates are used to display results with NBSview. Node coordinates are specified as a numeric array with three columns (x,y,z) and as many rows as there are network nodes. They can be given either as a:
    1. Valid Matlab expression (type the expression)
    2. Text file (*.txt)
    3. Matlab file (*.mat) containing a single variable
    Specifying node coordinates is optional, but they are mandatory for NBSview.
    See AALgetCOG.m for a script to determine the node coordinates (node centre of gravity) for an arbitrary node parcellation given as a NIFTI file.
    Use AALaalCOG.txt if the 116 AAL nodes are used.

      ④ 结点名称:

    NODE LABELS [optional]:
    Specify a label for each node. Node labels can be given either as a:
    1. Valid Matlab expression (type the expression)
    2. Text file (*.txt), where each line specifies a node label
    3. Matlab file (*.mat) containing a single variable
    The Matlab expression or the Matlab file must specify a cell array of strings, where there are as many strings as there are nodes.

    3)将数据添加至软件中:(① 在statistic test 处选择 t-test;② threshold 越大,留下的边对应的P值越小 【没有固定阈值推荐,尽量多试一些阈值】)

    4)结果展示:

    PS:如果有结果,软件会通过NBSview窗口自动弹出结果图。

    二、matlab 代码实现:

      1)使用nbsglm函数 ---> 主效应,交互效应;

      2)结合NBS软件中 nbs.m 找出其连通网络。

    参考连接:https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/GLM?highlight=%28GLM%29#Two_Groups_with_continuous_covariate_interaction

  • 相关阅读:
    spring mvc 总结
    linux安装tomcat及优化
    mysql支持emoji表情
    面试问题
    linux安装jdk mysql
    webstorm 介绍
    spring 总结
    UML工具
    js bom dom
    awt多线程聊天
  • 原文地址:https://www.cnblogs.com/hechangchun/p/13648172.html
Copyright © 2011-2022 走看看