US 9,811,906 B1
Method and data processing unit for segmenting an object in a medical image
Anamaria Vizitiu, Sfantu Gheorghe (RO); Olivier Ecabert, Ebermannstadt (DE); Jan Kretschmer, Nuremberg (DE); and Dominik Neumann, Erlangen (DE)
Assigned to SIEMENS HEALTHCARE GMBH, Erlangen (DE)
Filed by Siemens Healthcare GmbH, Erlangen (DE)
Filed on Apr. 26, 2017, as Appl. No. 15/497,453.
Claims priority of application No. 16179770 (EP), filed on Jul. 15, 2016.
Int. Cl. G06K 9/00 (2006.01); G06T 7/00 (2017.01); G06T 7/11 (2017.01); A61B 1/00 (2006.01)
CPC G06T 7/0012 (2013.01) [G06T 7/11 (2017.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 37 Claims
OG exemplary drawing
 
1. A method for segmenting an object in a medical image, the method comprising:
providing the medical image;
performing a plurality of iteration steps, input of a first iteration step of the plurality of iteration steps including the medical image as an input image and the input of each further iteration step of the plurality of iteration steps includes a vote map generated in a preceding iteration step as the input image, wherein each iteration step of the plurality of iteration steps comprises:
generating a plurality of patches, a portion of the input image and a patch location being assigned to each patch of the plurality of patches, the patch location being indicative of a location of the portion of the input image relative to the input image,
for each patch of the plurality of patches
determining a vote location based on the portion of the respective input image assigned to the respective patch and
determining a target location based on the vote location and the patch location assigned to that patch;
generating a vote map, wherein each patch of the plurality of patches contributes to a respective pixel value at the target location of the respective patch in the vote map;
determining a structure in the vote map of the final iteration step of the plurality of iteration steps;
determining a subset of patches of the plurality of patches of the first iteration step that contributed to the structure in the vote map of the final iteration step; and
segmenting the object based on the subset of patches to obtain a segmentation outcome.