t1 = neurobase::readnii("training01_01_t1.nii.gz")

Let’s do a marginal density of the values:

plot(density(t1)) # large spike at 0

You can also pass in a mask to most standard functions:

plot(density(t1, mask = t1 > 0))

Note the high intensities:

hist(t1)

The neurobase::ortho2 function displays nifti objects in 3 different planes.

neurobase::ortho2(t1)

We see a dark image; a this single large value affects how values are mapped. The function robust_window calculates Winsorizes an image, by default the 0 (min) and 99.9th quantile, and sets values outside of this range to that quantile (97.5th below).

ortho2(robust_window(t1, probs = c(0, 0.975)))

Note the x-axis:

For the rest of the slides we will use the robust t1 for plotting

Here we plot the T1 and a mask of values over \(300\):

ortho2(t1, y = t1 > 300)

Sometimes you would like to represent 2 images side by side, of the same dimensions and orientation of course (useful for checking registration), use double_ortho

double_ortho(t1, y = t1 > 300, col.y = "white")

The oro.nifti::image function shows a lightbox view, all slices of an image:

image(t1, useRaster = TRUE) # look at average brightness over each slice

The slice function can plot individual slices:

oro.nifti::slice(t1, z = c(60, 80))

We can specify z the same way but change the plane to be different to get a different slice of the brain (could also do coronal):

oro.nifti::slice(t1, z = 125, plane = "sagittal")

We can also overlay one slice of an image upon another using the oro.nifti::slice_overlay function.

slice_overlay(t1, y = t1 > 300, z = 80)

If you want to do 3D Gaussian smoothing, the extrantsr::smooth_image is helpful:

library(extrantsr)
sm_img = smooth_image(t1, sigma = 2)
double_ortho(t1, sm_img)

• ortho2 - show orthographic images (and with overlays)
• image - shows multiple slices of an image
• slice - shows only specified slices
• slice_overlay - similar to image but with an overlay
• double_ortho - similar to ortho2 but side-by-side
• robust_window - good for setting high values to not so high

http://johnmuschelli.com/imaging_in_r