Senologie - Zeitschrift für Mammadiagnostik und -therapie 2019; 16(02): e15
DOI: 10.1055/s-0039-1687978
Abstracts
Georg Thieme Verlag KG Stuttgart · New York

Speed-up of the reading workflow of combined x-ray and ultrasound breast images

M Hertel
1   Siemens Healthcare GmbH, Forchheim, Deutschland
2   Otto-Von-Guericke-Universität Magdeburg, Magdeburg, Deutschland
,
MW Beckmann
3   Universitätsklinikum Erlangen, Erlangen, Deutschland
,
P Fasching
3   Universitätsklinikum Erlangen, Erlangen, Deutschland
,
M Golatta
4   Universitätsklinikum Heidelberg, Heidelberg, Deutschland
,
M Juskic
4   Universitätsklinikum Heidelberg, Heidelberg, Deutschland
,
S Kappler
1   Siemens Healthcare GmbH, Forchheim, Deutschland
,
Q Li
1   Siemens Healthcare GmbH, Forchheim, Deutschland
,
C Liu
1   Siemens Healthcare GmbH, Forchheim, Deutschland
,
A Maier
5   Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Deutschland
,
R Nanke
1   Siemens Healthcare GmbH, Forchheim, Deutschland
,
G Rose
2   Otto-Von-Guericke-Universität Magdeburg, Magdeburg, Deutschland
,
B Schäfgen
4   Universitätsklinikum Heidelberg, Heidelberg, Deutschland
,
R Schulz-Wendtland
3   Universitätsklinikum Erlangen, Erlangen, Deutschland
,
E Wenkel
3   Universitätsklinikum Erlangen, Erlangen, Deutschland
,
M Radicke
1   Siemens Healthcare GmbH, Forchheim, Deutschland
› Author Affiliations
Further Information

Publication History

Publication Date:
28 May 2019 (online)

 
 

    Purpose:

    We tested an idea for a new reading software prototype capable of optimizing the current reading workflow of ultrasound in combination with x-ray mammography. Based on a small, non-representative study we measured the difference in workflow speed with and without using the software prototype.

    Materials and methods:

    A hardware prototype combining a 2-D mammography and a subsequent 3-D breast ultrasound scan was introduced earlier. In contrast to standard mammography (patient standing in upright position) and automated breast ultrasound (patient lying in horizontal position), the prototype acquires both images in the same breast geometry (upright position). Because mammography images are acquired in cone-beam geometry and the ultrasound images in parallel beam geometry, a misalignment up to 3.5 cm (geometrically calculated) must be overcome by scrolling through up to 70 slices in the ultrasound volume.

    Results:

    Instead of scrolling, we introduce a new, additional 2-D resliced plane which always contains the lesion. With only one additional selection the matching lesion will be automatically visualized in the original ultrasound volume and diagnosis can be performed in direct comparison. Six readers demonstrated thereby a speed-up of 74%± 22%.

    Discussions:

    To provide solutions to the rising demand of ultrasound diagnostics of mammographically dense breast patients, hardware prototypes (XUS (Siemens), McABUS (GE)) were introduced increasing the speed in the acquisition workflow. With the proposed software we targeted the reading workflow by contributing to higher workflow speed.


    #