Read Philips' EstimatedTotalReadoutTime and EstimatedEffectiveEchoSpacing

[oesteban]

Recent versions of dcm2niix have adopted these two for Philips scanners, we should check them before we resort to the water-fat shift formula.

See: rordenlab/dcm2niix#389 (rordenlab/dcm2niix@ab26453)

[oesteban]

@mgxd can you check if either EstimatedTotalReadoutTime or EstimatedEffectiveEchoSpacing are present in the metadata of those Philips scans we've got on Sherlock?

[mgxd]

No they are not - though if we had access to the DICOMs, we could try converting with the latest dcm2niix

[neurolabusc]

Philips DICOM images do not provide the required information. This is a limitation of Philips DICOM images, not dcm2niix. For details, see my comments here.

[manuegrx]

Hi,

This issue is quite old but, unless I missing something, it seems still relevant.
In my lab, we are using fMRIPrep for Philips data, and currently we add the field "TotalReadoutTime" in the json populated by dcm2niix in order to allow fMRIPrep to work for susceptibility distortion correction (using PEPOLAR techniques).

I was wondering if it was planned for fMRIPrep (and there for for sdcflows) to read the field "EstimatedTotalReadoutTime" when "TotalReadoutTime" is not available ?

Currently, it seems that fMRIPrep is not working without the field "TotalReadoutTime" because the bids validator returns an error (an issue is open on this subject : https://github.com/bids-standard/bids-validator/issues/1588) but, even without this error I am not sure that sdcflows will be able to read the field "EstimatedTotalReadoutTime"

[neurolabusc]

This is a limitation of the impoverished Philips DICOMs, not dcm2niix. Philips DICOMs are missing information required to determine BIDS tags like TotalReadoutTime. Please lobby the Philips clinical scientist associated your site to provide the required sequence details in the DICOM image files.