BIDS json file doesn't contain MultiBandAccelerationFactor for DWI sequences

[mvdoc]

Hi,
using heudiconv I realized that for DWI sequences dcm2niix doesn't store the field MultiBandAccelerationFactor. See nipy/heudiconv#110 for more info.

The information seem to be present in the dicom headers though (assuming I'm looking at the right field):

GRAPPA 2 (in this case the json file contains "ParallelReductionFactorInPlane": 2 correctly)

[bids@rolando test_dwi_dcm2niix] > grep 0051,1011 dump_dwi_grappa.txt
(0051,1011) LO [p2]                                     #   2, 1 Unknown Tag & Data

SMS 4

[bids@rolando test_dwi_dcm2niix] > grep 0051,1011 dump_dwi_sms.txt
(0051,1011) LO [s4]                                     #   2, 1 Unknown Tag & Data

This is the version we're using

bids@rolando:~/test_dwi_dcm2niix$ dcm2niix
Compression will be faster with 'pigz' installed
Chris Rorden's dcm2niiX version v1.0.20170923 (OpenJPEG build) GCC6.3.0 (64-bit Linux)

[neurolabusc]

I see you are using v1.0.20170923 - can you test to see if the latest release also has this problem. dcm2niix should read multiband directly from the CSA header rather than relying on the DICOM header. If there is still a problem, I would guess it is because you are using the Siemens SMS product rather than the CMRR sequences. I do not have the Siemens SMS license, so if this is the case I would be grateful if you could send me a sample dataset to examine.

[mvdoc]

I can confirm that with the latest release the MultiBand field is not reported. I'm emailing you a sample dataset of the DWI without acceleration/with grappa/with sms.

Thanks!

[neurolabusc]

Thanks for the example. If you compile the latest source code you will get v1.0.20171103 - this will detect multi-band from 0051,1011 if the CSA tag MosaicRefAcqTimes is missing. However, note that without MosaicRefAcqTimes you will not get slice timing parameters. The fact that MosaicRefAcqTimes is a defect of your images, and I suggest you work with your vendor (Siemens) to document and fix this issue with their VE11C.

[mvdoc]

Thank you!

[neurolabusc]

Hello, can I suggest you acquire your diffusion data with the mosaic option selected. I just checked our system, and if you do not have this checked each 2D slice is saved as a separate DICOM image and the MosaicRefAcqTimes CSA tag is omitted. Without that you will not get the SliceTiming field in your BIDS file. That will help you use eddy_cuda's --slspec file option. As an aside I keep the trace, ADC, FA and tensor buttons unchecked, as you can create better derived images after processing with dwidenoise, topup and eddy.

[mvdoc]

@neurolabusc thank you so much for this! I will try on our system and run some tests.

[mharms]

FYI: One peculiarity related to dMRI and the “Mosaic” output format is that at least under Siemens VE11 and VD13 (but I don’t think earlier versions), you have to set “Diff. weightings” >= 2 for the Mosaic option to be selectable. Assuming you are using a custom gradient table, we set “b-value 1” = 0, and “b-value 2” = <your desired maximum b-value>. You’ll then get an extra “b=0” volume at the front end of your acquisition, but you can use the Mosaic output format. Cheers, -MH

…

-- Michael Harms, Ph.D. ----------------------------------------------------------- Conte Center for the Neuroscience of Mental Disorders Washington University School of Medicine Department of Psychiatry, Box 8134 660 South Euclid Ave. Tel: 314-747-6173 St. Louis, MO 63110 Email: mharms@wustl.edu From: Chris Rorden <notifications@github.com> Reply-To: rordenlab/dcm2niix <reply@reply.github.com> Date: Thursday, November 9, 2017 at 1:42 PM To: rordenlab/dcm2niix <dcm2niix@noreply.github.com> Cc: Subscribed <subscribed@noreply.github.com> Subject: Re: [rordenlab/dcm2niix] BIDS json file doesn't contain MultiBandAccelerationFactor for DWI sequences (#141) Hello, can I suggest you acquire your diffusion data with the mosaic option selected. I just checked our system, and if you do not have this checked each 2D slice is saved as a separate DICOM image and the MosaicRefAcqTimes CSA tag is omitted. Without that you will not get the SliceTiming field in your BIDS file. That will help you use eddy_cuda's --slspec file option. As an aside I keep the trace, ADC, FA and tensor buttons unchecked, as you can create better derived images after processing with dwidenoise, topup and eddy. [mage removed by sender. mosaics]<https://user-images.githubusercontent.com/8930807/32624971-ed847b38-c558-11e7-9af2-24a63b81eb33.png> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#141 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ADswBoXh_7k012eMRI0Wyw6VrFgrKKc-ks5s01WegaJpZM4QQas8>.

________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

[mvdoc]

Today I tested with the mosaic option on, and everything works perfectly. Thank you!!