New script: dwicat

[Lestropie]

(name suggestions welcome)

As raised in multiple locations:

A convenience script for concatenating multiple DWI series where the intensity scaling of image reconstruction may have varied due to recalibration. This would involve:

• Extracting the b=0 volumes from each series.
  (Easiest would be to require .mifs at input so that gradient tables are embedded)

• Using mrhistmatch scale or mrstats -output median to determine the global intensity fluctuations between acquisitions.

• Concatenate all data into a single 4D image, applying appropriate multiplicative factors.

It should also check for variations in TE and TR between the input series, and warn where appropriate.

[Lestropie]

Bump

[jdtournier]

Sounds good, agree on the principle, and will definitely come in handy. I'd prefer the simpler name of dwicat personally. We could also make the intensity normalisation conditional on detecting a sufficiently large difference in scaling, along with a clear report about this? Also, I'd advocate using the raw DICOMs as input - though mifs would be fine too, of course.

[thijsdhollander]

Yep, 100% agreed with all of @jdtournier's comments.

[Lestropie]

Also, I'd advocate using the raw DICOMs as input - though mifs would be fine too, of course.

Any risk of interaction with dwidenoise?

[Lestropie]

Closed by #1626.