US 12,303,326 B2
Method of determining scan planes in the acquisition of ultrasound images and ultrasound system for the implementation of the method
Nicola Guraschi, Genoa (IT); Marco Crocco, Ovada (IT); and Luca Zini, Genoa (IT)
Assigned to Esaote S.p.A., Genoa (IT)
Filed by Esaote S.p.A., Genoa (IT)
Filed on Feb. 24, 2022, as Appl. No. 17/679,890.
Claims priority of application No. 102021000004376 (IT), filed on Feb. 25, 2021.
Prior Publication US 2022/0265242 A1, Aug. 25, 2022
Int. Cl. A61B 8/08 (2006.01); A61B 8/00 (2006.01); A61B 8/14 (2006.01)
CPC A61B 8/0866 (2013.01) [A61B 8/085 (2013.01); A61B 8/4245 (2013.01); A61B 8/469 (2013.01); A61B 8/483 (2013.01); A61B 8/5223 (2013.01); A61B 8/54 (2013.01); A61B 8/145 (2013.01)] 11 Claims
OG exemplary drawing
 
1. Method for determining scan planes in the acquisition of ultrasound images, comprising:
a) defining a body to be examined and defining a predetermined orientation and/or a predetermined position of a predetermined section plane which intersects said body at a desired location;
b) performing a three-dimensional volumetric ultrasound scan of said body by acquiring a plurality of scan planes via an ultrasound probe to acquire a volumetric ultrasound image comprising three-dimensional image data;
c) providing a machine learning algorithm for the identification of the image data from the volumetric ultrasound image that fall along said predetermined section plane that has the predetermined orientation and/or the predetermined position with respect to said body under examination and which said predetermined section plane intersects the volumetric ultrasound image, said machine learning algorithm being trained with a database of known cases to identify which data contributions of the image data from the volumetric ultrasound image fall on said predetermined section plane, or which of voxels that make up the image data of the volumetric ultrasound image fall on said predetermined section plane;
d) for each scan plane of the plurality of acquired scan planes, identifying the data contributions to the image data or the voxels that fall along the said predetermined section plane by means of the said machine learning algorithm according to step c) and generating therefrom a set of image data along the said predetermined section plane, said set of image data corresponding to a respective line of intersection between said predetermined section plane and each scan plane of the plurality of acquired scan planes;
e) storing said set of image data identified as falling along said predetermined section plane or subjecting said set of image data to further processing; and
f) determining, via the machine algorithm, the correspondence between a position and/or an orientation of a scan plane of the plurality of acquired scan planes with the predetermined position and/or predetermined orientation of the said predetermined section plane, and/or a correspondence between said set of image data related to the data contributions or the voxels that derive from or fall on the respective line of intersection between each scan plane of the plurality of acquired scan planes and the predetermined section plane.