US 11,810,287 B2
Defect detection in lyophilized drug products with convolutional neural networks
Zheng Li, South San Francisco, CA (US); and Calvin Tsay, Austin, TX (US)
Assigned to Genentech, Inc., South San Francisco, CA (US)
Filed by Genentech, Inc., South San Francisco, CA (US)
Filed on Jan. 7, 2022, as Appl. No. 17/571,399.
Application 17/571,399 is a continuation of application No. 16/654,200, filed on Oct. 16, 2019, granted, now 11,263,738.
Claims priority of provisional application 62/852,237, filed on May 23, 2019.
Claims priority of provisional application 62/748,304, filed on Oct. 19, 2018.
Prior Publication US 2022/0207708 A1, Jun. 30, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 7/00 (2017.01); G06N 20/00 (2019.01); G06N 3/04 (2023.01); G06F 16/53 (2019.01); G06F 17/15 (2006.01)
CPC G06T 7/0004 (2013.01) [G06F 16/53 (2019.01); G06F 17/15 (2013.01); G06N 3/04 (2013.01); G06N 20/00 (2019.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising, by a computing system:
receiving one or more querying images associated with a container of a pharmaceutical product, wherein each of the one or more querying images is based on a particular angle of the container of the pharmaceutical product;
calculating, for the container of the pharmaceutical product, one or more confidence scores associated with one or more defect indications, respectively, by processing a plurality of pixels of the one or more querying images using a target machine-learning model; and
determining, for the container of the pharmaceutical product, a defect indication from the one or more defect indications based on a comparison between the one or more confidence scores and one or more predefined threshold scores, respectively, wherein:
the defect indication comprises one or more of a critical defect, a major defect, or a minor defect,
the pharmaceutical product comprises a lyophilized product, and
the major defect comprises a cake appearance defect, the cake appearance defect comprising one or more of an extreme product splash, a slanted cake, or a shrunken cake.