Issue when converting data acquired with multiple slice groups

[jcohenadad]

Description on the issue

When the B1 mapping sequence based on a turboflash with a presaturation pulse (seq type, as shown on card: tfl, usual name in protocol: "tfl" or "presat_tfl") is acquired as interleaved with two slice groups and the the offset in the slice direction does not match the slice thickness, the resulting scans are not converted properly by DCM2NIIX. Conversion issues also persist when using SPM.

How to reproduce and test the issue

Acquire the interleaved B1 mapping sequence once with an offset between the two groups that matches the slice thickness and once where the offset is not matched.

Data: put it on gdrive
dcm2niix: v1.0.20210317
OS: Mac OS 10.15.7 (Catalina)
SPM version: SPM12 Version 7771

[dpapp86]

I converted all available data acquired with this sequence (acdc_7T_003p, 004p, 005 and 006, as well as Alan's shared data), and only found the issue once, when the offset did not match the slice resolution.

[dpapp86]

The only error was in acdc_spine_7T_005, during the only scan where slice thickness (2 mm) and offset (3mm) was not matched. DCM2NIIX conversion resulted in two files, one where the slice resolution was set to 3mm, and one where the slice resolution was set to 1mm.

[dpapp86]

SPM conversion always resulted the number of slices being increased by 1 (eg 41 slices instead of the 2x20 prescribed by the protocol).