US 12,257,070 B2
Pain assessment method based on deep learning model and analysis device
Sun Kwang Kim, Seoul (KR); Myeong Seong Bak, Seoul (KR); Hee Ra Yoon, Seoul (KR); Sang Jeong Kim, Seoul (KR); and Geehoon Chung, Seoul (KR)
Assigned to NEUROGRIN INC., Seoul (KR)
Filed by NEUROGRIN INC., Seoul (KR)
Filed on Jun. 22, 2022, as Appl. No. 17/846,772.
Application 17/846,772 is a continuation of application No. PCT/KR2020/006221, filed on May 12, 2020.
Claims priority of application No. 10-2019-0173382 (KR), filed on Dec. 23, 2019; and application No. 10-2020-0054412 (KR), filed on May 7, 2020.
Prior Publication US 2022/0323003 A1, Oct. 13, 2022
Int. Cl. A61B 5/00 (2006.01); G06T 7/11 (2017.01); G06V 40/16 (2022.01)
CPC A61B 5/4824 (2013.01) [A61B 5/7264 (2013.01); G06T 7/11 (2017.01); G06V 40/168 (2022.01)] 13 Claims
OG exemplary drawing
 
1. A method for pain assessment using a deep learning model, the method comprising:
receiving, by an analysis device, an image indicating activity in a specific brain area of a subject animal; and
allowing the analysis device to input images of regions of interest in the image into a neural network model and assess pain of the subject animal according to a result output by the neural network model,
wherein each region of interest is a region indicating the activity of an individual cell, and the analysis device inputs information generated based on the images of the regions of interest into a plurality of input layers of the neural network model, and
wherein the images are time-series data in a certain time interval, and the analysis device divides time-series data for the regions of interest into units of different sizes and inputs the time-series data divided into the units of different sizes into different input layers of the plurality of input layers.
 
9. A device for pain analysis using a deep learning model, the pain analysis device comprising:
an input device configured to receive an image indicating activity in a specific brain area of a subject animal;
a storage device configured to store a bidirectional recurrent neural network model that receives an image for brain activity and assesses a pain state; and
a computing device configured to input images of regions of interest in the image into a plurality of input layers of the bidirectional recurrent neural network model and assess pain of the subject animal according to a result output by the bidirectional recurrent neural network model,
wherein each region of interest is a region indicating the activity of an individual cell, and
wherein the images are time-series data in a certain time interval, and the computing device divides time-series data for the regions of interest into units of different sizes and inputs the time-series data divided into the units of different sizes into different input layers of the plurality of input layers.