Interactive live segmentation with automatic selection of optimal tomography slice

• US 9,269,141 B2
• Filed: 09/03/2012
• Issued: 02/23/2016
• Est. Priority Date: 09/07/2011
• Status: Active Grant

• Current Assignee: Koninklijke Philips N.V.
• Original Assignee: Koninklijke Philips N.V.
• Inventors: Wiemker, Rafael, Heese, Harald Sepp, Goossen, Andre, Erhard, Klaus
• Primary Examiner(s): Shah, Utpal
• Assistant Examiner(s): Zohrabyan, Narek
• Application Number:
  US14/342,553
• Publication Number:
  US 20140219534A1
• Time in Patent Office: 1,268 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06T 2200/24 involving graphical user interfaces [GUIs]
• CPC Class Codes: G06T 2207/20101 Interactive definition of point of interest, landmark or seed
• CPC Class Codes: G06T 2207/30068 Mammography; Breast
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/187 involving region growing; involving region merging; involving connected component labelling

Abstract

An apparatus for automatic selection of optimal tomography slices by executing a number of tentative segmentations using the same interactively provided in-slice seed point on some or all available tomography slices. The appropriate segmentation boundaries are then marked and the slice with the best segmentation goodness value (figure of merit) is presented to the user via a viewer. The steps are repeated when the user changes the seed point. The optimal segmentation boundary is displayed on top of a single simulated mammography image, fused from all tomography slice images.

Claims

1. An apparatus for post-processing 2D cross-sectional slice images, the apparatus comprising:

a processor, including;

a graphical user interface controller;

a segmenter unit; and

a goodness of segmentation evaluator;

wherein the graphical user interface controller is configured to generate for display on a screen a graphical user interface, the user interface when so displayed comprising an interactive first window widget displaying on the screen an initial slice based on one or more of the 2D slices, the window widget configured to allow a user to specify position data of a segmentation seed point anywhere on the displayed initial slice;

wherein the segmenter unit is configured to access, in response to the user specifying the position data, the 2D slices and to segment each of a plurality of the 2D slices by using the segmentation seed point as a common seed point for each of the segmentations;

wherein the goodness of segmentation evaluator is configured to establish from among the plurality of the segmented 2D slices a best slice having a segmentation with a best goodness of segmentation value; and

wherein the graphical user interface controller is further configured to update the first window widget to display therein, instead of the initial slice, the best slice or to display therein the best slice with the best segmentation overlaid or to generate for display on the screen in a second window widget the best segmentation overlaid on a second image based on all or a selection of the 2D slices.

2. The apparatus according to claim 1, wherein the second image is a simulated mammography image generated by forward projecting a 3D volume reconstructed from all or a selection of the 2D slices.

3. The apparatus according to claim 1, wherein the first window widget is updated by fade-in of the best slice.

4. The apparatus according to claim 1, wherein the position data of the common segmentation seed point is acquired by a user specifying with a pointer tool a position on the displayed first slice or on the displayed second image and, when the user specifies a new pointer position for a new common seed point, using the new common seed point to re-segment and re-establish a new best slice and a new best segmentation and then update the first window widget by displaying the new best slice instead of the previously displayed best slice or by displaying the new best slice with the new best segmentation overlaid or by overlaying the new best segmentation on the second image in the second window widget.

5. The apparatus according to claim 4, wherein the pointer tool is a computer mouse and wherein the specifying of the position data or the new position data is effected without using a mouse click event but by using position recordings generated by the mouse as the mouse is moved by the user.

6. The apparatus according to claim 1, wherein the processor comprises a back-tracer to back-trace the segmentation seed point across the plurality of the 2D slices by using a projection mapping to project the segmentation seed point into each one of the plurality of 2D slices thereby forming the common seed point, the back-tracer configured to read out metadata from the 2D slices and to change the projection mapping in dependence on the read out metadata.

7. A method of post processing 2D cross-sectional slice images, the method comprising:

displaying in a first window widget on a screen an initial image based on one or more of the 2D slices;

receiving position data of a segmentation seed point anywhere on the displayed initial image;

segmenting each of a plurality of the 2D slices by using the received seed point as a common seed point for each of the segmentations;

establishing from among the plurality of the segmented 2D slices a best slice having a segmentation with a best goodness of segmentation value; and

displaying on the screen in the first window widget the best slice or displaying in the first window widget the best slice with the best segmentation overlaid.

8. The method according to claim 7, further comprising;

displaying along with the first window widget a second window widget displaying therein an overview image having the best segmentation overlaid on same.

9. The method according to claim 7, wherein the segmenting into segments comprises:

receiving a plurality of sample points;

in each one of the plurality of 2D slices, solving across the plurality of the sample points a shortest-closed-path problem using a cost function inverse to a radial grey value gradient with respect to the common seed point, the common seed point being inside any one of the closed-paths; and

outputting a selection of the sample points as a solution to the shortest-closed-path problem, the outputted sample points making up a boundary of the respective segmentation.

10. The method according to claim 9, wherein the received plurality of the sample points are selected

i) to define as a whole a pre-determined mean radius around the common seed point;

and

ii) to yield in total at least a pre-determined mean grey value or mean grey value gradient.

11. The method according to claim 7, wherein the establishing of the best goodness of segmentation value comprises:

for at least two slices, calculating for each a goodness of segmentation value for the segments in each one of the at least two slices, the calculating step comprising;

along a or the segmentation boundary, summing the projection of the gray value gradient in the slice; and

/or

along the segmentation boundary, establishing the mean of the projection of the grey value gradient in the slice; and

after calculating the goodness of segmentation value,

comparing the values to establish a best segmentation value; and

selecting the slice having the best or a better segmentation value.

12. Method according to claim 7, wherein the 2D slices are digital breast tomosynthesis (DBT) slices.

13. A non-transitory computer readable storage medium having stored thereon a computer program element for controlling an apparatus which, when executed by a processing unit, is adapted to perform the steps according to claim 7.