Why is the implementation of Enhanced DICOM delayed

Why is the implementation of Enhanced DICOM delayed ? Kees Verduin Chair DICOM WG 16 www. kees-verduin. nl 2

Presentation outline Reasons for Delayed Implementations Benefits of Enhanced SOP Classes The clinical examples of its benefits for MRI Problem areas Suggestions for Solution 3

Delayed ? Facts: Enhanced MR SOP class in DICOM standard since 2002 Demonstrations by Modality Vendors in 2004 and 2005 Expectations: Implementations by 2006 with MR and CT modalities Massive support by PACS and Workstations Substantial support through IHE profiles All these are Delayed ! 4

What happened till today? Two modality vendors actually implemented the creation of Enhanced MR objects. Many PACS vendors support the Storage only. A few PACS/Workstation vendors do support DISPLAY of more than just the pixels Late support by IHE (till 2009), however, these IHE profiles are still under debate because of interpretation/implementation differences New attributes are also added to the classic IODs. 5

Reason … DICOM is non-committal. When approving a supplement, nobody in the industry commits itself for implementation. Therefore good idea’s may lay around for years and vendor debates may go on for years. Users can only address the vendors individually. 6

Why is there still a debate ? �The Enhanced DICOM standard for MR and CT supports many FEATURES, without ONE OVERWHELMING BENEFIT. �There is no direct competitive advantage thus: No Business Priority �During implementation vendors found that the architecture of their systems does not simply allow the support by their own workstations and PACS systems and therefore delayed the implementation. 7

But vendors in DICOM agreed? �The DICOM standard has become too diverse to simply implement everything. �The DICOM representatives are NOT responsible for implementation and prioritization. �Prioritization: “the art of choosing what NOT to do”, where one choice is : “do nothing and await what happens”. �DICOM Standards Committee does not look at implementation speed. 8

But there were such successful demos at SCAR and RSNA ? Yes, these were stimulated AND executed by development staff from several vendors. However, only those who were also responsible for product policy, could steer towards implementation. 9

What are these Benefits ? Its NEW, supporting State of the Art imaging It solves existing problems with Private Attributes It is compact, through a Multi-Frame header It adds Real World Values Its adds Color (also in full RGB) It also supports storage of Spectroscopy It also supports storage of Raw Data 10

A new MR Standard (Supplement 49) Dimensions Multi-frame Spectroscopy Color Multi-stack Raw Data Real World Values 11

Diffusion Imaging “Diffusion b-values” sorted (and an ADC image) (courtesy: Philips) 12

Perfusion Imaging non perfused stroke area time Stor ed valu es Real World Value Slope (0040, 9225) Signal Real World Value Intercept (0040, 9224) RW values delayed perfusion Quantitative data with Real World Values (courtesy: Siemens) time-to-peak map 13

Functional Brain Imaging � 10 -60 slices �all slices measured in one TR �repeated 100 -1000 times to get sufficient signal �leading to > 60, 000 images in one object Store thousands of images in one object and display them in a consistent way using Multi-frame Header and Dimension Module 14

Spectroscopy and its Imaging Relative NAA peak-height Ratio of Choline and Creatinine peaks 15

Problems …. There was a timing dilemma foreseen for changing from “classic” MR and CT to Enhanced DICOM This dilemma is still an actual problem 16

The timing dilemma for the Enhanced MR object For the MR vendors: Why implement it, while nobody is ready to use it? For the Workstation vendors: Why implement it, when no one is creating it? the PACS vendors: For Support STORAGE, but how to deal with the mix of workstations that can yes/no receive it ? 17

The timing dilemma for the Enhanced MR object If nobody is asking for it , nobody will solve the chicken and egg problem Only the PACS/Workstation vendors can solve this. 18

Problems and adversaries Current discussions in WG 16 are about (potential) implementation choices by different vendors like: Concatenations Dimension Organization Object selection Series definition 19

How can this delay be addressed ? More buy-in from modality vendors: better cooperation in the interest of the end-users More involvement by end-users/organizations: more clinical focus/influence for IHE Profiles More involvement of Workstation/PACS vendors: to solve the chicken and egg problem All stakeholders should be represented in/by the Committee for the Advancement of DICOM 25

What should be done now ? Power to the users: Join IHE user groups, Demand improvements from your vendors, Require up-to-date DICOM implementation, Join DICOM working groups. More Power to the DICOM Standards Committee: Demand timely implementation, Reduce non-committal approval of new supplements. 27

Thank you for your attention Muito Obrigado 29

Acknowledgement and copyright The slides of this presentation (including the hidden slides) may be quoted with reference to the author : www. kees-verduin. nl email: kees@kees-verduin. nl 30

31