US 12,300,006 B2
Method and system for digital staining of microscopy images using deep learning
Aydogan Ozcan, Los Angeles, CA (US); Yair Rivenson, Los Angeles, CA (US); Hongda Wang, Los Angeles, CA (US); Yilin Luo, Los Angeles, CA (US); Kevin de Haan, Los Angeles, CA (US); Yijie Zhang, Los Angeles, CA (US); and Bijie Bai, Los Angeles, CA (US)
Assigned to THE REGENTS OF THE UNIVERSITY OF CALIFORNIA, Oakland, CA (US)
Appl. No. 17/783,260
Filed by THE REGENTS OF THE UNIVERSITY OF CALIFORNIA, Oakland, CA (US)
PCT Filed Dec. 22, 2020, PCT No. PCT/US2020/066708
§ 371(c)(1), (2) Date Jun. 7, 2022,
PCT Pub. No. WO2021/133847, PCT Pub. Date Jul. 1, 2021.
Claims priority of provisional application 63/058,329, filed on Jul. 29, 2020.
Claims priority of provisional application 62/952,964, filed on Dec. 23, 2019.
Prior Publication US 2023/0030424 A1, Feb. 2, 2023
Int. Cl. G06V 20/00 (2022.01); G06T 7/00 (2017.01); G06V 10/82 (2022.01); G06V 20/69 (2022.01)
CPC G06V 20/698 (2022.01) [G06T 7/0012 (2013.01); G06V 10/82 (2022.01); G06T 2207/10056 (2013.01); G06T 2207/20084 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A method of generating a virtually stained microscopic image of a test sample comprising:
providing a neural network that is executed by image processing software using one or more processors of a computing device, wherein the neural network is trained with a plurality of pairs of stained and matched microscopic images or image patches, wherein each pair comprises a first image or image patch of a training sample that is virtually or chemically stained with a first stain type and a second image or image patch of the training sample that is virtually or chemically stained with a second, different stain type;
obtaining an input image of the test sample that is virtually or chemically stained with the first stain type;
inputting the input image of the test sample into the neural network; and
transforming, via the neural network, the input image into an output image of the test sample that is virtually stained with the second stain type to substantially resemble and match a stained microscopic image of the test sample that is chemically stained with the second stain type.