Run `sct_compute_compression` only once

[valosekj]

This PR updates the processing script for dcm-zurich to be compatible with recent updates of sct_compute_compression.

Context: sct_compute_compression now outputs both normalized and non-normalized values into a single file; moreover, values from individual subjects are aggregated into a single file; see https://github.com/spinalcordtoolbox/spinalcordtoolbox/pull/4003#discussion_r1246007753 and https://github.com/spinalcordtoolbox/spinalcordtoolbox/pull/4172

TODO:

• [x] wait https://github.com/spinalcordtoolbox/spinalcordtoolbox/pull/4172 is merged, then run the script on dcm-zurich

Fixes: https://github.com/sct-pipeline/dcm-metric-normalization/issues/17

[valosekj]

Processing finished without errors. Results copied to ~/duke/projects/dcm-normalization/dcm-zurich_2023-08-13_persex. The metrics across all subjects are now aggregated in a single CSV file (~/duke/projects/dcm-normalization/dcm-zurich_2023-08-13_persex/results/compression_metrics.csv). This is a difference from the previous analysis (dcm-zurich_2023-05-17_persex), where each metric for each subject was saved in a separate CSV file. A single CSV should be easier to handle in further analyses --> we can simplify the read_MSCC function in statistics/compute_stats_zurich.py.

@sandrinebedard, if you agree, we can merge the PR.