US 11,769,594 B2
Deep learning model learning device and method for cancer region
Won Tae Kim, Suwon-si (KR); Shin Uk Kang, Seoul (KR); Myung Jae Lee, Seoul (KR); Dong Min Kim, Seoul (KR); and Jin Seong Jang, Seoul (KR)
Assigned to JLK INC., Cheongju-si (KR)
Appl. No. 17/284,778
Filed by JLK INC., Cheongju-si (KR)
PCT Filed Oct. 11, 2019, PCT No. PCT/KR2019/013398
§ 371(c)(1), (2) Date Apr. 12, 2021,
PCT Pub. No. WO2020/076135, PCT Pub. Date Apr. 16, 2020.
Claims priority of application No. 10-2018-0121393 (KR), filed on Oct. 11, 2018; and application No. 10-2018-0121513 (KR), filed on Oct. 12, 2018.
Prior Publication US 2021/0391078 A1, Dec. 16, 2021
Int. Cl. G06K 9/00 (2022.01); G16H 50/20 (2018.01); G16H 30/20 (2018.01); A61B 5/055 (2006.01); A61B 5/00 (2006.01); G06N 3/08 (2023.01); G06T 7/00 (2017.01)
CPC G16H 50/20 (2018.01) [A61B 5/055 (2013.01); A61B 5/7267 (2013.01); G06N 3/08 (2013.01); G06T 7/0014 (2013.01); G16H 30/20 (2018.01); A61B 2576/00 (2013.01); G06T 2207/10016 (2013.01); G06T 2207/10088 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30081 (2013.01)] 6 Claims
OG exemplary drawing
 
1. A deep learning model learning device for a cancer region, the deep learning model learning device comprising:
a parametric MRI image input part for inputting an image corresponding to a diagnosis region, inputting at least one parametric MRI (Magnetic Resonance Imaging) image constructed on the basis of parameters different from each other, and constructing and providing an MRI moving image by using the at least one parametric MRI image;
a cancer detection model learning part for receiving an input of the at least one parametric MRI image and the MRI moving image corresponding to the diagnosis region, and learning a deep learning model on the basis of information labeling the cancer region;
a labeling reference information providing part for providing at least one reference information contributing to the labeling of the cancer region; and
a labeling processing part for checking the cancer region input on the basis of the at least one reference information and processing the labeling of the checked cancer region,
wherein the cancer detection model learning part comprises an image-based MRI learning part that constructs a base layer of the deep learning model through learning on the at least one parametric MRI image,
the cancer detection model learning part comprises a learning model correction part that constructs an extension layer of the deep learning model through learning on the MRI moving image, and
the learning model correction part comprises:
a first learning model correction part for constructing a first extension layer of the deep learning model through learning on a first MRI moving image; and
a second learning model correction part for constructing a second extension layer of the deep learning model through learning on a second MRI moving image.