A utility to put the DICOM standard into a searchable database with web app
This uses python to parse a local copy of the DocBook xml version of the standard (see DICOM Standard status page for the most up-to-date version: http://www.dclunie.com/dicom-status/status.html#BaseStandard2013).
The paragraphs are converted into couchdb documents and pushed to the server. Text search is enabled by Lucene index over paragraph text. As such, this implementation, as is, is tied to the specific instance of Cloudant database.
A utility couchSite copies the local site directory as attachments to a document called .site so that the site can be hosted directly from CouchDB. The site allows you to type a keyword and get instant results.
couchdb - https://pypi.python.org/pypi/CouchDB
lxml - https://pypi.python.org/pypi/lxml/3.4.4
Some things that this doesn't support:
-
searches for words less than 5 letters long
-
boolean operations or wildcards
-
figures, tables, and other items from the standard
-
a DICOM data dictionary for quick lookup
-
search in the titles of the DocBook links (tables, sections, etc.)
Development of this search index was supported in part by the Quantitative Image Informatics in Cancer Research (QIICR) project (http://qiicr.org) through the award U24 CA180918 from the National Cancer Institute.