US 12,462,386 B2
Autonomous cell imaging and modeling system
Hervé Marie-Nelly, San Francisco, CA (US); Jeevaa Velayutham, Shah Alam (MY); Zachary Phillips, San Francisco, CA (US); and Shengjiang Tu, Foster City, CA (US)
Assigned to Insitro, Inc., South San Francisco, CA (US)
Filed by Insitro, Inc., South San Francisco, CA (US)
Filed on Dec. 6, 2024, as Appl. No. 18/972,720.
Application 18/972,720 is a division of application No. 18/527,037, filed on Dec. 1, 2023, granted, now 12,198,344.
Application 18/527,037 is a division of application No. 18/111,405, filed on Feb. 17, 2023, granted, now 11,875,506, issued on Jan. 16, 2024.
Application 18/111,405 is a continuation of application No. PCT/US2022/080200, filed on Nov. 19, 2022.
Claims priority of provisional application 63/281,536, filed on Nov. 19, 2021.
Prior Publication US 2025/0095151 A1, Mar. 20, 2025
Int. Cl. G06T 7/00 (2017.01); A61B 5/00 (2006.01); G01N 15/1429 (2024.01); G16H 20/00 (2018.01)
CPC G06T 7/0012 (2013.01) [A61B 5/4848 (2013.01); G01N 15/1429 (2013.01); G16H 20/00 (2018.01); G06T 2207/10064 (2013.01); G06T 2207/20036 (2013.01); G06T 2207/30024 (2013.01)] 23 Claims
OG exemplary drawing
 
1. A system for identifying one or more cell culture conditions for progressing one or more live biological cells towards a desired cell state, the system comprising:
one or more processors;
a memory; and
one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors to:
perform a plurality of cell culture condition identification cycles on the one or more live biological cells until a condition is met, wherein each cell culture condition identification cycle comprises:
obtaining a set of one or more images capturing the one or more live biological cells;
determining, from at least a subset of the set of one or more images, a cell state of the one or more live biological cells;
identifying a new cell culture condition for progressing the one or more live biological cells towards the desired cell state, by inputting the cell state and the desired cell state into a trained machine-learning model, wherein the trained machine-learning model is an active-learning machine-learning model;
applying the new cell culture condition to the one or more live biological cells;
prompting a user to provide one or more user inputs about a state of the one or more live biological cells after the new cell culture condition is applied; and
retraining the active-learning machine-learning model based on the user inputs; and
identify the one or more cell culture conditions, based on outcomes of the plurality of cell culture condition identification cycles, for use in progressing live biological cells towards the desired cell state in future cell cultures.