DICOM WG 22 AND AMERICAN DENTAL ASSOCIATION WORKING GROUP 12.1: DICOM MEETING American Dental Association Chicago, Illinois July 11-12, 2005
MEMBERS PRESENT: Allan G. Farman (WG 22/12.1 – Univ. Louisville/ADA Representative to DICOM) Chris Bope (Soredex - TUUSULA Finland) Ari Kontkanen (Planmeca USA) Tony Bavuso (Progeny) Hang Yuan (Progeny) Kelly Anderson (Dentrix) Filippo Impieri (Kavo/Gendex – Des Plaines, IL) Sam Marfatia (Kavo/Gendex – Des Plaines, IL) Xavier Carayol (Kodak Imaging – Paris, France) Manny Pena (Schick Technologies, Long Island City, NY) Andrew Casertano (VA) John Goyette (Schick/WG 22/WG 12.1) Greg Zeller (Veterans Affairs) Don Vu (J. Morita, USA) Lance Guffey (Patterson Dental) Jason Wendling (Patterson Dental) Eriks Blaschka (Sirona Dental) Jolanta Majewska (ADSTRA Systems, Toronto, Canada) Jim Pearson (AAO) Dan Croley (ADA) Scott Benjamin (AIM - WG 12.4) Steve Jambor GUESTS: Paul Bralower, American Dental Association, Chicago, IL
1. Introduction of a new ADA IT Manager – Jim Pearon 2. Paul Bralower update on status of TR 1023. Ballot closes on Thursday, July 14, 2005 3. Overview of WG 12.1, WG22 and TR1023 on its scope, work items and goals to the first time attendees presented by Allan Forman 4. Minutes approval from February 22-23 delayed to afternoon after a printed copy is available 5. ADA interoperability project overview of the past years and the usage of the CDs outside ADA trade show presented by John Goyette 6. Final validation of DICOM images will be done on Friday, July 08, 2005 by David Clunie Timeline for the CD project from last meeting, Feb. 24-25, 2005 Review errors from David’s last validation o Max file length is 8 characters; must be rename on FTP site o Retired Person Name o Review of the DICOM image file Patient Orientation is not correct in some of the images: should be L not P. Dataset in the image doesn’t correspond to the actual file. Kelly Anderson (Dentrix) explanation of the tag – Patient Orientation: L is direction of row and F is direction of columns 7. Overview of ADA CD Datasets problems: Un-enhanced images must be submitted for the CD Orientation of the image is critical Datasets problems per vendor: ADSTRA: Schick images pixels wrongly inverted PA Orientation (wrong tooth number) CEPH image should have Patient Orientation A/F not L/F DENTRIX: PAN was encoded as IO DEXIS: Image Orientation needs to be corrected GE:
8.
Patient Name tag is different in each image CEPH Image Pixel Spacing is zero and Patient Orientation should be A/F not F/A
Modality wrong for PAN (IO) only in DICOMDIR – Hermut Feuerhahn to fix this bug
KODAK: PAN image should have Modality PX not DX Pixel Spacing on CEPH is wrong IO image should be labeled 36 FDI and Patient Orientation should be P/F PATTERSON: IO AirTechninques images Pixel Spacing is wrong (24.5) PLANMECA: CEPH images Patient Orientation should be A/F not L/F PROGENY Patient Name incorrect; it should be Company and Product Name Re-label IO patient Laterality, Image Orientation and Tooth Number (L, P/F, ###) not (R/F, R, ###): up to the vendor to specify There is nothing wrong with those images SIRONA TOMO/PAN image missing tags L R on the image itself. Allan Farman comments on the TOMO image; Sirona will replace the TOMO image since WG 22 did not address that issue yet Sirona IO Bitewing image has wrong region. It should be jaw region and only one tooth number instead of nine teeth numbers TIGERVIEW Tags should be relabeled CEPH wrong Patient Orientation PAN wrong Image Laterality and Patient Orientation GENDEX IO Sensor image: Image Laterality is wrong; should not be B CEPH should be A/F not L/F for Patient Orientation DenOptix image Modifier Sub Sequence is wrong level Patient First Name is a different one on each image (Visualix, DenOptix IO) Discussion: 9. Missing info describing the way the images are displayed in different software. Image dataset is correct and has all the required info. Action: Members agreed that all the images will be displayed from the outside-in. The software right now does not use the orientation correctly.
For the CD we need to re-label the teeth on some of the pictures as per Allan Farman to keep consistency. 10. CD content overview Any image changes need to be submitted to FTP site; any suggestions to Power Point Presentation by August 8 th Any changes to Logos should be sent to Allan Farman Checks should be sent to Paul Bralower (ADA) by August 1, 2005
11. Overview of the Power Point Presentation from last year’s CD presented by Allan Farman. Manufacturer asked to comment and to check if there are any changes to the name or missing logos and provide them to Allan Farman. 13. E-mail Attachment Supplement: Work item proposal forwarded by Xavier Carayol Recommended: Proposal 1: new ZIP Interchange Profile - Zip DICOM images into one file Proposal 2: add to Standards new media type DICOM ZIP Memo to be submitted to International Internet Committee to use DCZ extension in order for the application be able to open the DICOM image and DICOMDIR automatically. Action: 1st step would be to submit the new DCZ extension. The security issues are outside of this scope. A request for a Work Item has already been written, motioned and passed. Any comments on this Work Item are welcomed. Xavier Carayol will write TR describing all stated in paragraph 13 by October 4 th, 2005 meeting. VA – Dr. Greg Zeller presents Templates: VA Imaging Status Update to DICOM WG-22 VistA Imaging overview Annex A and Radiology Core document Vendor conformance: VA will begin Alpha Testing of DICOM with Q/R VA/DoD Annex: any feedback welcomed VA needs the info of when the study was created, when the image is captured (Acquisition Date) and when the file is actually created (Content Date) Action: Add to Feb. 24-25, 2005 minutes a VA/DoD Joint DICOM Req. document and comments for TR1023. 14. 1. 2. 3. 4. 5. 6.
15.
John Goyette overview/update of current work item: Dental Viewsets
There is no consistent way to display the viewsets. This is why we need to develop a standard method of encoding and exchanging view sets. New IO might be required. Hanging Protocols are approved but have limitation. Hanging Protocols are not associated with Patient/Study/Series/Image. There is no clear way to handle custom orientations and larger number of images. It’s very complex. Proposal: Develop new IOD Timeline: March 2006 Finished September 2005 = 1 st draft available November 2005 = 1 st Group 6 reading (standard committee to approve this) Dental Viewset Object:
a. Patient, Study, Series IODs b. Frame of reference for the display area i. i.e. (0, 0) to (1, 1) ii. Physical Dimension (mm) c. Sequence of views or view boxes i. Special coordinates Upper Left (0.3, 0.3) to Lower Right (0.4, 0.06) w/h ii. Reference to contained image 1. Study UID 2. Series UID 3. Image UID d. Z-Order e. Aspect Ratio Deleted i. BEST FIT BOX ii. BEST FIT IMAGE f. Preferred Acquisition Context i. Tooth Numbering ii. Order Sequence iii. Orientation of the sensor iv. (Sensor Size) g. Annotation Overlay (Text and Graphics) h. Color? – Brent/Allan – user definable not needed here
16.
VA viewsets discussion:
a. User search for FMX by Name, Description of viewsets and number of images. Store info for retrieval of mounts so they will be displayed the same way as the referring Doctor’s view.
b. FMX will be displayed in the following order, starting top-left going to right then bottom across to left and up. This way the images are kept the in order of time when the X-Rays where taken c. When the viewset is received and rearranged, it will become a Hanging Protocol d. Viewset is an instance of Hanging Protocol (template) unique to a Patient e. Viewset object explains how to display it f. Viewset is like a hyperlink to the image g. Viewset object does not contain image, it will only contain a list of images UIDs. VA sends it in consistent order all of the time h. We need to use at higher level. What’s the expected Orientation of the Images? i. Scott suggested sending images which are modified (Presentation stage). A derived image is to be sent to the other person j. VA implementation of DICOM to exchange viewsets based on consistent order. This is what VA is using:
Modality Tag Attribute Name
IO - Intra Oral
(0008,0020) (0008,0030) (0008, 0023) (0008, 0033) (0008,0060) (0008,2218) (0008,2220)
Study Date Study Time Content Date Content Time Modality Anatomic Region Sequence Anatomic Region Modifier Sequence
Primary Anatomic Structure Sequence (0008,2228) (0020,0062) (0020, 0020) (0008,2218) (0020,0062) (0020,4000) (0008,2218) Image Laterality Patient Orientation Anatomic Region Sequence Image Laterality Image Comments Anatomic Region Sequence
PX – Panoramic DX – Cephalometric XC – Visible Light Photographic
Modality
Tag
Attribute Name
(0008,2220)
> Anatomic Region Modifier Sequence
Primary Anatomic Structure Sequence (0008,2228) (0020,0062) (0020,4000) Image Laterality Image Comments
ES – Visible Light Endoscopic
k. VA/DoD document section 2.6 “Sorting Presentation Constructs” explains how the viewsets could be sorted 17. John Goyette will be leading Dental Viewsets Work Item in one team, together with Kelly Anderson (Dentrix), Eriks Blaschka (Sirona) to write technical aspects and attributes. Second team: Allan G. Farman, Andrew Casertano (VA), Greg Zeller (VA) and Scott Benjamin to write USE CASES. Deadline: October 4, 2005. Action: Approve minutes from the last meeting. No comments. Motion was seconded and passed. New Item: Multi-Frame of DX images storage specification proposal by Xavier Carayol. Defining DX Multi-frame Images objects (e.g. could be used for implants) not currently supported by the Standard. The goal of this proposal is to enable an application to save, in a single DICOM Part 10 type of file, a set of Digital X-Ray images created during the same tomography acquisition process or by a specific post-processing application program in order to simplify such image file manipulation. New DX Object based on existing one Enhanced DX Image Module with added two modules. Multi Frame Functional Group Remove some unused Modules. Modify some DX Patient Orientation tags to have V, P, M (medial) and D (distal). Since the image is 3D the anatomical direction is useful. Working Group 17 attempted to develop n-D standards but it was dropped. A J.Morita image has one hundred 2D frames which are put into one 3D image.
Sets of slices will be specified as viewset. Pixel Measures Macros contains info about pixel spacing and the slice and thickness. DX Frame Content Macro. Recommendation: Xavier would like to have a proposal for September meeting to present to DICOM. This document will be posted on FTP and any comments are welcome, especially from other CT vendors. Discussion: 1. Modality: any type 2. Image Type: it would be useful in this case to differentiate CT 3D image from any other 3. Separate Modality proposal might be rejected by DICOM gurus. John Goyette suggested TMJ, TOMO and Sinus slices could be put in a new TAG 4. Scott Benjamin’s suggestion was to have a separate class since it will be used for oral dental 5. Xavier proposed something very generic and uses a simple Identifier to get Modality for this purpose 6. Steve Jambor proposed to keep it as strictly dental and everyone agreed 7. Patient Orientation position 2D is not enough which is why 3 rd value is necessary here Action: An initial draft of work item request to be written by Xavier Carayol to get input from other manufacturers. Proposed completion date: Dec 2005. John Goyette motioned to put this as a work item. Motion was seconded and approved. Visible Light by Scott Benjamin: ADA WG12.4 TR1024. PatientID CD project. Feedback welcomed from vendor for the case below: 1. Select images and click on one button to export and burn them to the CD. The tags and image format are not important. The CD will then be used for different purposes (e.g. child abduction). Image requires anatomy and patient orientation, also contain xrays, if they are present in DB. 2. Pre-defined Report generator: It would be some sort of batch file; specialized report feature to be added to software. TR 1029 talks about different views of the image and with six different attributes the image will be described. Meeting Adjourned.
FUTURE MEETINGS Oct 4th, 2005 next meeting in Philadelphia. Submitted by Jolanta Majewska (ADSTRA Systems Inc.) Reporting Secretary
Approved:
__________________________ Allan G. Farman, BDS Co-Chair
_________________________ John Goyette Co-Chair
Reviewed by Counsel
December 21, 2005