US 11,734,824 B2
Image processing method, drug sensitivity test method and image processing apparatus
Tamio Mizukami, Shiga (JP); and Katsumi Kishimoto, Shiga (JP)
Assigned to FRONTIER PHARMA INC., Shiga (JP)
Appl. No. 17/59,500
Filed by FRONTIER PHARMA INC., Shiga (JP)
PCT Filed May 17, 2019, PCT No. PCT/JP2019/019666
§ 371(c)(1), (2) Date Nov. 30, 2020,
PCT Pub. No. WO2019/230447, PCT Pub. Date Dec. 5, 2019.
Claims priority of application No. 2018-105777 (JP), filed on Jun. 1, 2018; and application No. 2019-089529 (JP), filed on May 10, 2019.
Prior Publication US 2021/0224992 A1, Jul. 22, 2021
Int. Cl. G06K 9/00 (2022.01); G06T 7/00 (2017.01); G06T 7/73 (2017.01); G16H 20/10 (2018.01); G16H 30/40 (2018.01); G06N 20/00 (2019.01)
CPC G06T 7/0014 (2013.01) [G06N 20/00 (2019.01); G06T 7/74 (2017.01); G16H 20/10 (2018.01); G16H 30/40 (2018.01); G06T 2207/10064 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30024 (2013.01)] 10 Claims
OG exemplary drawing
 
1. An image processing method for detecting a position of a detected part which is specific from a test image obtained by imaging a cell, the image processing method comprising:
generating an image which is corresponding to the test image and in which a pseudo marker is attached as an intermediate image by inputting the test image in which a marker corresponding to the detected part is not expressed to a first learning model and;
generating an image in which the position of the detected part is indicated by a representative point thereof by inputting the intermediate image to a second learning model and outputting the image as a result image, wherein:
the first learning model is constructed by using teacher data associating a first image and a second image captured to include a same cell and performing deep learning with the second image corresponding to an input and the first image corresponding to an output, the first image being an image in which the marker is expressed and the second image being an image in which the marker is not expressed; and
the second learning model is constructed by using teacher data associating a third image, which is captured to include a cell and in which the marker is expressed, and information representing a position of the representative point included in the third image and performing deep learning with the third image corresponding to an input and the position of the representative point corresponding to an output.