US 12,444,052 B2
Learning apparatus, learning method, program, trained model, and endoscope system
Seiya Takenouchi, Tokyo (JP)
Assigned to FUJIFILM Corporation, Tokyo (JP)
Filed by FUJIFILM Corporation, Tokyo (JP)
Filed on Feb. 8, 2023, as Appl. No. 18/165,935.
Application 18/165,935 is a continuation of application No. PCT/JP2021/027440, filed on Jul. 26, 2021.
Claims priority of application No. 2020-144569 (JP), filed on Aug. 28, 2020.
Prior Publication US 2023/0206445 A1, Jun. 29, 2023
Int. Cl. G06K 9/00 (2022.01); A61B 1/00 (2006.01); G06T 7/00 (2017.01); G06V 10/774 (2022.01); G06V 10/82 (2022.01); G16H 10/40 (2018.01)
CPC G06T 7/0014 (2013.01) [A61B 1/000096 (2022.02); G06V 10/774 (2022.01); G06V 10/82 (2022.01); G16H 10/40 (2018.01); G06T 2207/10068 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30096 (2013.01); G06V 2201/03 (2022.01)] 11 Claims
OG exemplary drawing
 
1. A learning apparatus comprising:
a recognizer implemented by a neural network; and
a processor,
the processor being configured to:
acquire a learning image and biopsy information associated with the learning image, the learning image being obtained by capturing an image of an examination object, the biopsy information being information indicating a location where a biopsy for the examination object has been performed;
generate ground-truth region data in which a region including the location where the biopsy has been performed is set as a ground-truth region, on the basis of the biopsy information, wherein generate the ground-truth region data comprising:
generate the ground-truth region data by pairing a biopsy information image which includes the location of the biopsy with a candidate ground-truth region data image; and
train the recognizer that recognizes a region of interest, by using the learning image and the ground-truth region data.