I have encountered both formats using different test scripts. This makes me very confused. Can the organizer confirm which format will eventually be used? This will eliminate **one degree of freedom** from our system design so that is a good thing. Thanks!

Created by Li Shen thefaculty
Hello, The Tensorflow example expects .dcm files. = Darvin
Thanks Li! @darvinyi Hi Darvin, Can you check that the Tensorflow example expects ".dcm" files instead of "dcm.gz"? Thanks!
Hi Thomas, When I tested the tensorflow example code, it didn't work because the inputs were "*.dcm.gz". I had to modify the code to make it work. But that was like 1-2 weeks ago. In my recent tests, it seems all files are "*.dcm". So this does not appear to be a problem anymore. I just want to confirm it. Thanks!
Hi Li, It's not clear to me whether you are reporting that you have found .dcm.gz in /trainingData or if you are asking if you should expect to find .dcm.gz there. I don't find any occurrences of ".dcm.gz" in the recent documentation ([Submitting Models](https://www.synapse.org/#!Synapse:syn4224222/wiki/401759)). The only occurrences that come to my mind is that we provide the pilot images under this format, obviously because it doesn't make sense to make you download them uncompressed (.dcm). If you are reporting the presence of .dcm.gz files in /trainingData, can you please show a file listing (`find /trainingData -name "*.dcm.gz"`)? If you have found an occurrence of ".dcm.gz" in the documentation, can you please give the reference to this page? Thanks!

.dcm or .dcm.gz files under the /trainingData folder? page is loading…