User Tools

Site Tools


neuroimagen:byicv

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
neuroimagen:byicv [2015/01/28 09:53]
osotolongo created
neuroimagen:byicv [2020/08/04 10:58] (current)
Line 1: Line 1:
-====== How to correct by Intracraneal Volume (ICV) ======+====== How to correct by Intracranial Volume (ICV) ====== 
 +{{ :neuroimagen:adjusted.png?nolink&300|}}
  
-I'm going to suppose you did the pipeline thing over your MRI images and you already have the /study_mri.csv/ file. +I'm going to suppose you did the pipeline thing over your MRI images and you already have the //study_mri.csv// file. The Freesurfer version I'm using makes an estimation of Intracraneal volume and the pipeline scripts put it as //EstimatedTotalIntraCranialVol_eTIV// into the results file. First do this for a real CSV file: 
 +<code bash> 
 +$ sed 's/;/,/g' ab255_mri.csv > ab255_mri_rc.csv 
 +</code> 
 + 
 +Let's say we want to get the hippocampus volume (HV) corrected by ICV. Then we must get Left and Right HV and calculate its medium value. **HV = 0.5*(HVR+HVL)** 
 + 
 +Now first step is to fit a straight line for HV //vs// ICV and the ICV mean value (<ICV>). The adjusted HV will be **aHV = HV - b*(ICV-<ICV>)**, with //b// as the regression coefficient of HV //vs// ICV. 
 + 
 +Now we need to get the data. Let's try it in R. 
 +<code R> 
 +x <- read.csv("ab255_mri_rc.csv"
 +x$HV = 0.5*(x$"Right.Hippocampus"+x$"Left.Hippocampus"
 +a <- lm(x$HV~x$"EstimatedTotalIntraCranialVol_eTIV"
 +b=a$coefficients[[2]] 
 +x$aHV = x$HV - b*(x$EstimatedTotalIntraCranialVol_eTIV - mean(x$EstimatedTotalIntraCranialVol_eTIV, na.rm=TRUE)) 
 +write.csv(x, file="ab255_mri_aHV.csv"
 +</code> 
 + 
 +==== For ADNIMERGE package ==== 
 + 
 +{{ :neuroimagen:adni_hippocampus_correction.png?300|}} 
 + 
 +El paquete ADNIMERGE provee los valores de Hipoocampo (suma de ambos) e ICV. Puede corregirse directamente y dividirse entre 2 para obtener valores similares. 
 + 
 +<code R> 
 +library("ADNIMERGE"
 +a <- lm(adnimerge$Hippocampus~adnimerge$ICV) 
 +adnimerge$adj.Hippocampus = 0.5*(adnimerge$Hippocampus - a$coefficients[[2]]*(adnimerge$ICV - mean(adnimerge$ICV, na.rm=T))) 
 +</code> 
 + 
 +Este procedimiento se debera aplicar para realizar la replica de cualquier estudio donde se utilize la semisuma corregida de ambos hipocampos.
neuroimagen/byicv.1422438793.txt.gz · Last modified: 2020/08/04 10:46 (external edit)