This function helps create scatter plots with marginal histograms (we call this plot a scattergram) to help assess the threshold grid and fit of the TAPAS model.
The data contained in this package contains 3D arrays for 15 synthetic gold standard segmentations (gs#), 15 probability maps (pmap#), and a single brain mask (brain_mask) that can be used for all 30 images. The # value noted can be any number 1-15. Package data is set to Lazy Data: true and will therefore be automatically loaded with the package. We use 10 subjects for training the TAPAS model and then evaluate the model using 5 subjects excluded from training in this tutorial. The gold standard segmentations and probability maps consist of a single slice with voxel values and a second slice where all voxels are equal to 0. The brain mask contains a brain mask value in both slices. These are saved as arrays which we will convert to nifti objects. The authors would like to re-iterate that all data available in the package and used throughout this vignette are synthetic and created by the authors simply for demonstration of the package usage. The data is meant to mimic a simple set of real data.