1 / 37

The Radiology Playbook: A Missing Component of RadLex™

The Radiology Playbook: A Missing Component of RadLex™ . Nasya Mendoza-Elias and Lizhou Huang Pat Mongkolwat, Ph.D. and David S. Channin, M.D. Northwestern University Feinberg School of Medicine DePaul University College of Computing and Digital Media

liam
Download Presentation

The Radiology Playbook: A Missing Component of RadLex™

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. The Radiology Playbook:A Missing Component of RadLex™ Nasya Mendoza-Elias and Lizhou Huang Pat Mongkolwat, Ph.D. and David S. Channin, M.D. Northwestern University Feinberg School of Medicine DePaul University College of Computing and Digital Media National Science Foundation Research Experience for Undergraduates In Medical Informatics Program August 21, 2008

  2. Overview • Introduction • Progress • Future Work

  3. What is RadLex™? • Funded by the Radiological Society of North America (RSNA) • Purpose • Capturing • Indexing • Retrieving • Will unify other lexicons

  4. What is the RadLex™Playbook? • Orders, procedures, and procedure steps • Powered by an ontology • Dynamic

  5. Why do we need a Playbook? • Consistency • Interoperability • Inadequacy of other lexicons

  6. What aboutCommon Procedural Terminology CPT™? Not an adequate description • 71010 Radiologic examination, chest; single view, frontal • 71015 Radiologic examination, chest; stereo, frontal • 71020 Radiologic examination, chest, two views, frontal and lateral • 71021 Radiologic examination, chest, two views, frontal and lateral; with apical lordotic • 71022 Radiologic examination, chest, two views, frontal and lateral; with oblique • 71023 Radiologic examination, chest, two views, frontal and lateral; with fluoroscopy • 71030 Radiologic examination, chest, complete, minimum of four views • 71034 Radiologic examination, chest, complete, minimum of four views; with fluoroscopy • 71035 Radiologic examination, chest, special views (eg, lateral decubitus, Bucky studies)

  7. Whereas Merrill’s Atlas of Radiographic Positioning and Radiologic Procedures says:

  8. The IHE Model • Order Placer System: orderable • Order Filler: orderable procedures • Order Filler: procedure steps via modality worklist

  9. So… • Orderable  Referring physician • Requested Procedure(s)  Radiologist • Reportable, codifiable, billable unit of work • Procedure Step  Technologist (modality) • Scheduled procedure steps • Performed procedure steps

  10. The Challenge Treading the fine line of granularity between the needs of the referring physician and the detail of radiology operations

  11. What will the RadLex™ Playbook deliver initially? • Knowledge model • Grammar • Set of orderable names • Grammar for imaging and non-imaging procedure steps • A set of procedure step names • Can be used in DICOM

  12. What can you do with the Playbook? • Build a chargemaster • Protocol studies • Tailor studies • Build software applications with “knowledge”

  13. What’s been keeping us busy • Ontology structure • GUI • Query • Filtering of choices

  14. Ontologies • What is an ontology? • Concepts and relationships • For example • Wine: An alcoholic beverage fermented from grapes • Red Wine: Wine fermented from whole grapes • is_a wine • color: red • White wine: Wine fermented from skinless grapes • is_a wine • color: white • Chardonnay is_a white wine • Bordeaux is_a red wine • Allows for knowledge based computation: • “Is Chardonnay a wine?”

  15. For example… NMH-HEAD_47 • MR FACE MP SCOUT • MR FACE SAG T1R-2D(SE_1)_N • MR FACE COR T1R-2D(SE_1)_N • MR FACE COR T2P-2D(SE_N)_N/M • MR FACE AX T1R-2D(SE_1)_N • MR FACE AX T2P-2D(SE_N)_N/M • MR FACE INJ GAD • MR FACE AX 2D(GRE_1)_N • MR FACE COR 2D(GRE_1)_N • MR FACE SAG 2D(GRE_1)_N Paranasal Sinus • Survey • SAG T1 SE • COR SE T1 • COR T2 TSE • AX SE T1 • AX T2 TSE • INJECT • AX SHARP FS • COR SHARP FS • SAG SHARP FS

  16. Translation –CT, DEXA, Fluoro, Mammo, X-Ray • Use RadLex • Categorize procedure steps by attributes

  17. Knowledge Model Example • XRAY CHEST is_a Imaging Orderable • XRAY CHEST has_steps • XRAY CHEST PA • XRAY CHEST PA is_a XRAY Procedure Step • XRAY CHEST PA has_projection CORONAL • XRAY CHEST PA produces DICOM DX Object • XRAY CHEST LATERAL LEFT • XRAY CHEST LATERAL LEFT is_a XRAY Procedure Step • XRAY CHEST LATERAL LEFT has_projection SAGGITAL • XRAY CHEST LATERAL LEFT produces DICOM DX Object

  18. 1 2

  19. 4 5 3

  20. 7 6 10 8 11 9

  21. 12

  22. Name Rest. 1 Rest. 2 Rest. 3 Rest. 4 Rest. 5

  23. Procedure Step

  24. User Interface • Query based on a series of restrictions • Restriction filters are mutually linked • Restrictors can be applied in any order

  25. GUI (July 18)

  26. 1 2

  27. 3 4

  28. 5

  29. Querying the Ontology • Code by Kaustubh Supekar, Stanford, created for a different project • Surprisingly roundabout code • a.txt • Deletes everything that is not a result

  30. Problems with the query • Slow • Requires vast amounts of code • Loads ontology for every query

  31. Future Work • Connecting all parts of the GUI with the query • Optimizing the query • Strip unnecessary code • Load ontology once

  32. Future Work • Realtime web interface • Modifying ontology • Authorization and administration • Integration into RadLex • Other features

  33. Such as… • Exporting protocols • Printing • Searching for similar protocols

  34. Thank you Questions?

More Related