Add dicom as import

[khelm]

This adds a complete list of DICOM tags as datatype properties as well as some terms extracted from the DICOM specification docs. It also makes a number of other fixes: 1) In some DICOM tag description, references are made to pairs of values in the format ## or sets of values like "column/row". Unfortunately, the "\" is considered by rdflib as an escape character, so if the text is "-1-1", rdf lib is thinking of this as "-". So I changed the "\"'s to "/" which is not an escape character.

The following also use "\", now replaced by "/". The context is usually "column\row" or "##" dicom:dicom_0072003C dicom:dicom_00181310 dicom:dicom_00700014 dicom:dicom_00480202 dicom:dicom_00700011 dicom:dicom_00700010 dicom:dicom_00460202 dicom:dicom_00700053 dicom:dicom_00700052 dicom:dicom_00700022 dicom:dicom_00420014 dicom:dicom_00286114 dicom:dicom_00460213 dicom:dicom_60xx0050 dicom:dicom_00221468 dicom:dicom_00221467 dicom:dicom_00720038 dicom:dicom_00480201 dicom:dicom_00480008 dicom:dicom_00181620 dicom:dicom_00289503 @cmaumet and @natachaperez: Also, there are unicode characters in the definitions and, in at least one case, in what is assigned as a key in OwlReader. This still throws an error but it's not from the OWL file, but rather from the OwlReader class in owl_reader.py. Line 52, where it casts the key as a string - str(key) - is the current error. Fix by making the cast: unicode(key) in /home/Karl/Work/packages/miniconda2/envs/nidm-specs/lib/python2.7/site-packages/nidmresults/owl/owl_reader.py which gets called in owl_to_webpage.py This fixes the issue.

Added new self.owl.graph.bind statements for some of the imported vocabularies (dct, dicom, pato, prov, qibo). Already existing in owl_to_webpage.py were (nidm, onli, and sio).

[natachaperez]

@khelm ++1 for your job with dicom terms !