INVALID classificaiton should not prevent basic cataloging

[hjmjohnson]

Describe the bug:
When creating a catalog of dicom files to write to a json file, if a file that is invalid to process for classificaiton is encountered, an entry in the json file sould still be created.

To Reproduce:

Steps to reproduce the behavior:

python3 ~/Botimageai/dcm-classifier/scripts/classify_study.py -d /nas4321/Private_Reference_Data/TEST_DCM_CLASSIFIER_FAILURE  -j test.json

Expected behavior:
Any valiid dicom series should be represented in the json file, even if it can not be explicitly classified.

In place of the "GUESSED*" dictionary elements, it sould be nice to have the reson why a classification could not be made.

The "StudyInstanceUID" must be listed, along with all other dicom elements.

Actual Behavior:

An empty dictionary is created . (See series 700 which I think is a screen shot).