Timothy Hendrickson

Another comment on the particular DICOM data that I am dealing with is that identifying information such acquisition and study date has been stripped, so having a feature that allows...

As a feature, returning `None` if something doesn't exist makes sense to me.

@yarikoptic, let me speak to the owners of the data whether there are any data use certifications/authentications at play. I'll be in touch!

@yarikoptic, I was able to get some phantom data from the same scanner and with the same de-identification routines applied. Attached are DICOM folders representing three localizers. Let me know...

Hi @yarikoptic, just checking in on the status of this. Best, -Tim

The data that we have is de-identified so it is entirely likely that no dates will be present within the DICOM header, as those represent PHI. Is it doable to...

Just checking up on this. Best, -Tim

@pvelasco Thanks for getting back to me. Yes, the DICOMs are organized in the way you suspected. Here is an example: `20211001-ST001-SUB1000101ses1/MR-SE030-rest_ep2d_bold_ME_16MIN_PhysioLog: MR-ST001-SE030-0001.dcm 20211001-ST001-SUB1000101ses1/MR-SE031-rest_ep2d_bold_ME_16MIN2_SBRef: MR-ST001-SE031-0001.dcm MR-ST001-SE031-EC3-0001.dcm MR-ST001-SE031-EC5-0001.dcm MR-ST001-SE031-EC2-0001.dcm MR-ST001-SE031-EC4-0001.dcm `...

Thanks for the suggestion. I tried upgrading heudiconv to 0.5.3 and got another error: **Traceback (most recent call last): File "/usr/bin/heudiconv", line 11, in load_entry_point('heudiconv==0.5.3', 'console_scripts', 'heudiconv')() File "/usr/lib/python2.7/dist-packages/heudiconv/cli/run.py", line...

Downgrading torchmetrics to `v0.6.0` seems to resolve the issue.