Make a custom parquet file for TCIA

[fedorov]

See details in aregesteruams/NBIA_data_retriever_gui#6

[jmwhorton]

      Here is the current list of columns we are using in the parquet file and would need in the extract:

       SeriesInstanceUID
       series_aws_url
       PatientID
       StudyInstanceUID
       collection_id
       series_size_MB
       StudyDate
       StudyDescription
       SeriesNumber
       SeriesDescription