Breast Tomosynthesis - Brief Proposal

1. Proposed Workitem: Breast Tomosynthesis

Proposal Editor: David Clunie
Editor: TBD
Date: N/A (Wiki keeps history)
Version: N/A (Wiki keeps history)
Domain: Radiology

2. The Problem

Breast Tomosynthesis (Digital Breast Tomography (DBT)) is a new modality that shares many features with digital mammography, but has additional specific requirements to assure interoperability between vendors as well as satisfaction of user's functional requirements for display.

3. Key Use Case

A patient undergoes repeat screening mammography with tomosynthesis, and comparison is needed with last year's prior exam, which may have been either ordinary digital mammography or may have involved tomosynthesis, with possibly a different vendor or model of acquisition modality, possibly at a different institution.

The interpreting radiologist requires that the current and priors be displayed with the necessary repertoire of mammography display features in the IHE MAMMO Profile, with the additional features required to properly display tomosynthesis reconstructed slices, synthetic views, and if necessary, tomosynthesis projection images and/or CAD, without resorting to the need for an acquisition-vendor-specific proprietary workstation.

4. Standards and Systems

IHE MAMMO
IHE MAWF
DICOM Breast Tomosynthesis SOP Class
DICOM Sup 165 Breast Projection X-Ray Image Storage SOP Class (in * development)

5. Discussion

The IHE MAMMO profile has been one of the most effective IHE activities in terms of influencing the standardization of widespread support for the idiosyncratic requirements of mammography in general purpose PACS and viewers.

Breast tomosynthesis is in need of the same form of attention, and now is the time to act, since penetration of the user base is limited but increasing and users are demanding that their enterprise infrastructure support tomosynthesis rather than keeping it segregated in dedicated systems. Very soon, the problem of incompatible outside priors also will become apparent, if standards are not adopted.

The primary concern is the communication and display of reconstructed slices, and any synthetic views displayed from them, but standard support of projection images, and of CAD, is also desirable.

There is a small risk that DICOM Sup 165 will not be completed in time, in which case the projection image aspect could be deferred.

CAD encoding for Tomosynthesis is not yet defined, though could theoretically be by the time this profile is completed; worst case, requirement of CAD display could also be deferred, though it has proven to be a very important feature of the existing IHE MAMMO profile.