Hi, I have multiband dwi data acquired with a Siemens scanner and I successfully converted most of these data using dcm2niix version v1.0.20211006. However, a sizable amount of participants did not convert correctly- I received the error message in the attached screenshot. 

Does this suggest an error in data collection or conversion? If the problem lies within the conversion process, what do you suggest I do to address it?

I appreciate your time and assistance!

Jen

This suggests that some of the DICOM images are missing - e.g. not copied by the PACS, or disk space was exhausted. In rare situations this suggests a reconstruction failure. One option is to go to the MRI console and export the data directly to a USB drive. You may also want to look at the sequence and see how many volumes/slices were expected. What Siemens system are you using (e.g. VE11, XA30, etc.). If you are using an XA system, I would suggest exporting the data as enhanced format instead of these classic images. 

If there is no way to recover the missing images, you could sort your images by instance number and remove the partial volumes. For your data you would run 
 dcm2niix -r y /path/to/DICOMs
this will create a renamed set of DICOMs where the filename starts with the instance number, you would want to discard all images except the initial contiguous volumes. This is likely to result in an unbalanced DWI sampling, but it may be better than having no data.

This is also a great opportunity to leverage your Siemens Research Collaboration Manager. Their role is to help you resolve situations like this.

Thank you for your quick response! The data were indeed missing DICOMS and are in the process of being recovered. Thank you!