Add statistics for PMJ-based CSA and compute distance from C2-C3 and PMJ

[sandrinebedard]

Description

This PR addresses #83 and #84. Main changes include:

In process_data.sh:

• Compute distance from PMJ along extrapolated centerline from C2-C3 disc with the script get_distance_pmj_disc.py, save in a .csv file

To to this, I added the saving of the centerline coordinates since it is more precise than in the image of centerline_extrapolated --> https://github.com/spinalcordtoolbox/spinalcordtoolbox/pull/3478/commits/79752f834c9cb54b6a8b244068f1556a61dd8f75

In get_subject_info.py:

• Read PMJ-based CSA and a column for it
• Read distance from PMJ and C2-C3 disc and add a column for it

In compute_stats.py:

• Compute mean, std, median, COV for both CSA measures
• Compute regression, normalization only with PMJ-based CSA
• Add scatter plot for C2-C3-based CSA vs PMJ-based CSA
• Add scatter plot for distance from PMJ and C2-C3 disc

Linked issues

Resolves #84 and Resolves #83

[sandrinebedard]

I should maybe add in the ylabelof figure a) that CSA is measured at 60mm from PMJ

[sandrinebedard]

I just have a question: what is the distance from the PMJ and how did you choose it?

Currently it is 60 mm!

Now that I have the distance between PMJ and C2-C3 disc, the mean is at 63 mm, but those results don't include yet the new R-L adjustment of PMJ label, so it should be arround that! I want to compute PMJ at the mean distance from C2-C3 disc