In the words of the RSNA
"As images, imaging reports, and medical records move online, radiologists need a unified language to organize and retrieve them. Radiologists currently use a variety of terminologies and standards, but no single lexicon serves all of their needs. RSNA RadLex is a single unified source of radiology terms that is designed to fill this need.
Beginning in 2005, RSNA convened experts in imaging informatics and radiological subspecialties to create this resource, which is now made freely available to the healthcare community. RadLex has developed into a rich, structured radiology-specific ontology, which currently includes more than 30,000 terms."
More of the detail
Broadly speaking, it is separated into dependent and independent constituents. This may not appear to be an obvious initial separation, but is based on the underlying idea that a term (or "entity") may exist in a state that is independent from another (e.g. anatomy, imaging acquisition) or be dependent on an "independent entity", e.g. pathology, procedure, imaging observation.
One of the most interesting parts of the ontology is the pathology section which details the relationships between pathology as experienced in clinical practice.
Top level headings
- dependent continuant
- pathophysiologic process
- procedure step
- Radlex descriptor
- imaging observation
- independent continuant
- medical object
- report component
- anatomical entity
- imaging acquisition
- foreign body
- imaging modality
That's all well and good, but what earthly good is such a system? Well, the proposal is that it will be helpful in many situations:
- automatic order entry decision support
- vendor independent "protocoling" of complex imaging exams
- reliable PACS display layouts
- no need to re-dictate lengthy imaging technique sequences
- improved speech recognition accuracy
- speech-enabled structured reporting to satisfy regulatory requirements
- real-time decision support for the radiologist
- rapid teaching file creation
- accurate report search and data mining
Some of this may be specific to the Americas, but there will be a worldwide interest in such a system.
- 1. Langlotz CP. RadLex: a new method for indexing online educational materials. Radiographics. 26 (6): 1595-7. doi:10.1148/rg.266065168 - Pubmed citation
- 2. Hazen R, Van esbroeck AP, Mongkolwat P et-al. Automatic Extraction of Concepts to Extend RadLex. J Digit Imaging. 2011;24 (1): 165-9. doi:10.1007/s10278-010-9334-1 - Pubmed citation
- 3. Rubin DL, Flanders A, Kim W et-al. Ontology-assisted analysis of web queries to determine the knowledge radiologists seek. J Digit Imaging. 2011;24 (1): 160-4. doi:10.1007/s10278-010-9289-2 - Pubmed citation
- 4. Liu K, Chapman WW, Savova G et-al. Effectiveness of Lexico-syntactic Pattern Matching for Ontology Enrichment with Clinical Documents. Methods Inf Med. 2010;49 (6)doi:10.3414/ME10-01-0020 - Pubmed citation
- 5. Do BH, Wu A, Biswal S et-al. Informatics in radiology: RADTF: a semantic search-enabled, natural language processor-generated radiology teaching file. Radiographics. 2010;30 (7): 2039-48. doi:10.1148/rg.307105083 - Pubmed citation
- 6. Kundu S, Itkin M, Gervais DA et-al. The IR Radlex Project: an interventional radiology lexicon--a collaborative project of the Radiological Society of North America and the Society of Interventional Radiology. J Vasc Interv Radiol. 2009;20 (7 Suppl): S275-7. doi:10.1016/j.jvir.2009.04.031 - Pubmed citation