Vertebral levels vs vertebral bodies

In https://github.com/ivadomed/model-spinal-rootlets/pull/53, we use intervertebral discs to compute the distance between each disc and the PMJ (script [here](https://github.com/ivadomed/model-spinal-rootlets/blob/ffbde356ba7a0d695ff9a24b0b73db16482a5b43/pediatric_rootlets/vertebrae_to_spinal_levels.py)). These distances are taken as proxies for the vertebral **_levels_** and plotted along with spinal levels using [this script](https://github.com/ivadomed/model-spinal-rootlets/blob/ffbde356ba7a0d695ff9a24b0b73db16482a5b43/pediatric_rootlets/generate_figure_rootlets_and_vertebral_spinal_levels.py). This approach results into a continuous visualization of the vertebral levels:

(legend: red - spinal levels, green - vertebral levels)

An alternative approach could be to vizualize vertebral _**bodies**_ as done in [Cadotte et al., 2015](https://pubmed.ncbi.nlm.nih.gov/25523587/) (shown in light yellow):

Vertebral bodies could be obtained, for example, using SpineNet ([repo](https://github.com/rwindsor1/SpineNet/tree/main), [tutorial](https://github.com/rwindsor1/SpineNet/blob/main/tutorials/03-spinegeneric-example-detections.ipynb), [paper](https://www.nature.com/articles/s41598-024-64580-w)):


Start with pediatric_rootlets/vertebrae_to_spinal_levels.py and pediatric_rootlets/generate_figure_rootlets_and_vertebral_spinal_levels.py, then review PR #53 to understand the current disc-based measurements and visualization. Compare this approach with the proposed vertebral-body visualization using the linked SpineNet tutorial and determine which representation the project should implement.

Written by the indexing model from the issue text.