Hello MricroGL users,

While I was converting my multi echo bold data to nifti I receive the following warnings:

Found 600 DICOM file(s)
Slices not stacked: echo varies (TE 14, 46.16; echo 1, 3). Use 'merge 2D slices' option to force stacking
Warning: Adjusting for negative MosaicRefAcqTimes (issue 271).
UINT16->INT16 Future release will change default. github.com/rordenlab/dcm2niix/issues/338
Convert 200 DICOM as /Volumes/rsfMRI_RC/RS_Data/sub-027/sub-027_session2/func/sub-027_ses-2_rest_bold_e1 (64x64x43x200)
Warning: Adjusting for negative MosaicRefAcqTimes (issue 271).
UINT16->INT16 Future release will change default. github.com/rordenlab/dcm2niix/issues/338
Convert 200 DICOM as /Volumes/rsfMRI_RC/RS_Data/sub-027/sub-027_session2/func/sub-027_ses-2_rest_bold_e3 (64x64x43x200)
Warning: Adjusting for negative MosaicRefAcqTimes (issue 271).
UINT16->INT16 Future release will change default. github.com/rordenlab/dcm2niix/issues/338
Convert 200 DICOM as /Volumes/rsfMRI_RC/RS_Data/sub-027/sub-027_session2/func/sub-027_ses-2_rest_bold_e2 (64x64x43x200)
Conversion required 12.313125 seconds (5.550481 for core code).

Do I need to be worried about the outputs? It looks like MricroGL does convert my bold data into 3 separate echos plus the BIDS sidecars (json) and place them into the chosen folder snd I do not see any issues. But "adjusting for negative mosaicrefacq times" warning got me worried bit??
In advanced setting I only chose 1)ignore derived and 2D images 2)precise Philips scalling 

Has anybody had the same issue or successfully converted multi echo data with mricrogl before? I would appreciate any pearls of wisdom or reassurance.

Thank you,

Remy

I would make sure to include the dcm2niix version number in your report, and ensure it is the latest stable version (v1.0.20210317). It sounds like everything has processed correctly. I would check the slice times in the JSON files. I wonder if this data was acquired with Siemens MoCo, which can do odd things to slice times (my own preference is to acquire data without motion correction, and model motion effects in your subsequent analyses).

Thank you for your response Chris, and apologies for the late reply, I just saw this. I checked the scanner protocol pdf (Siemens Magnotom Prisma_Fit) and  acquisitions were done without motion correction, so that is not the issue and my dcm2niix version is same as above. I ended up using Horos to separate the echoes and then the brik_head AFNI method to prep my dataset for ME-ICA. I am going to give it a try one more time as I want the data to be available publicly in BIDS format, but in the meantime if you came up with a solution, I would appreciate if you could share it here. thank you again!

I suspect this is simply a warning message. You will want to make sure the SliceTiming field in the JSON matches your acquisition.

I am going to check, and if they did not match I will let you know. Thank you!

Also, I could share the data with you! I can share it via cloud (Microsoft one note) If I have your email address or you could email me at rcohan@yorku.ca!

Your images look fine. The warnings will not appear in the next release of dcm2niix, This was addressed with the 5th May commit:

Reduce false alarms for warning 'Adjusting for negative MosaicRefAcqTimes

https://github.com/rordenlab/dcm2niix/co...

Chris, thank you very much for looking into this!