US 12,257,041 B2
Method, device, and computer program for predicting brain tissue lesion distribution
Oh Young Bang, Seoul (KR); Yoon-Chul Kim, Seoul (KR); Jong-Won Chung, Seoul (KR); Woo-Keun Seo, Seoul (KR); Gyeong-Moon Kim, Seoul (KR); Geon Ha Kim, Gwacheon si (KR); and Pyoung Jeon, Seoul (KR)
Assigned to Samsung Life Public Welfare Foundation, Seoul (KR)
Filed by Samsung Life Public Welfare Foundation, Seoul (KR)
Filed on Nov. 16, 2020, as Appl. No. 17/099,008.
Claims priority of application No. 10-2020-0075985 (KR), filed on Jun. 22, 2020.
Prior Publication US 2021/0393212 A1, Dec. 23, 2021
Int. Cl. A61B 5/055 (2006.01); A61B 5/00 (2006.01); A61B 6/50 (2024.01); G06F 18/214 (2023.01); G06F 18/22 (2023.01); G06T 7/00 (2017.01); G06V 10/25 (2022.01); G06V 10/50 (2022.01); G06V 10/75 (2022.01); G06V 10/80 (2022.01); G06V 10/82 (2022.01); G16H 30/40 (2018.01); G16H 50/20 (2018.01); G16H 50/50 (2018.01)
CPC A61B 5/055 (2013.01) [A61B 5/0042 (2013.01); A61B 5/7267 (2013.01); A61B 5/7275 (2013.01); G06F 18/214 (2023.01); G06F 18/22 (2023.01); G06T 7/0014 (2013.01); G16H 30/40 (2018.01); G16H 50/20 (2018.01); G16H 50/50 (2018.01); A61B 6/501 (2013.01); A61B 2576/026 (2013.01); G06T 2207/20076 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30016 (2013.01); G06T 2207/30096 (2013.01); G06V 10/25 (2022.01); G06V 10/50 (2022.01); G06V 10/758 (2022.01); G06V 10/811 (2022.01); G06V 10/82 (2022.01); G06V 2201/031 (2022.01)] 9 Claims
OG exemplary drawing
 
1. A method of predicting a brain tissue lesion distribution, the method comprising:
a model learning operation of learning a prediction model for predicting a brain tissue lesion distribution in a subject by using brain image data of a plurality of previous patients, wherein the model learning operation includes an image matching operation of matching different types of the brain image data of each of the plurality of previous patients, an operation of calculating deformation data from the brain image data and selecting a region of interest corresponding to the lesion site, an operation of labeling the region of interest with a preset value according to whether or not each voxel corresponds to a lesion and a learning input obtaining operation of obtaining learning input data by extracting the deformation data for each voxel with respect to the region of interest;
an input obtaining operation of obtaining input data from brain image data of the subject;
an input operation of inputting the input data into the prediction model; and
an output operation of generating output image data including information on the lesion distribution after recanalization treatment for the subject, by using the prediction model,
wherein the prediction model includes a success prediction model that is learned using pre-treatment and post-treatment brain image data of a plurality of first patients who have modified Treatment In Cerebral Infarction (mTICI) scores of 2b and 3 among the plurality of previous patients, and a failure prediction model that is learned using pre-treatment and post-treatment brain image data of a plurality of second patients who have mTICI scores of 0, 1 and 2a among the plurality of previous patients,
wherein the brain image data matched in the image matching operation includes first diffusion weighted image (DWI) data obtained before treatment, perfusion weighted image (PWI) data before treatment, and second diffusion weighted image (DWI) data obtained after treatment, and
wherein the calculating and selecting operation includes
an operation of calculating an apparent diffusion coefficient (ADC) from the first diffusion weighted image (DWI) data;
an operation of calculating relative time to peak (rTTP) from the perfusion weighted image (PWI) data before treatment to obtain an rTTP map; and
an operation of selecting the region of interest from the first diffusion weighted image (DWI) data, the perfusion weighted image (PWI) data before treatment, and the second diffusion weighted image (DWI) data.