US 12,170,142 B2
Classification based on characterization analysis methods and systems
Mustafa Jaber, Los Angeles, CA (US); Liudmila A Beziaeva, Culver City, CA (US); Christopher W Szeto, Scotts Valley, CA (US); and Bing Song, La Canada, CA (US)
Assigned to NantOmics, LLC, Culver City, CA (US); and NantHealth, Inc., Culver City, CA (US)
Filed by NantOmics, LLC, Culver City, CA (US); and NantHealth, Inc., Culver City, CA (US)
Filed on Apr. 26, 2023, as Appl. No. 18/139,550.
Application 18/139,550 is a continuation of application No. 17/539,292, filed on Dec. 1, 2021, granted, now 11,676,707.
Application 17/539,292 is a continuation of application No. 16/685,191, filed on Nov. 15, 2019, granted, now 11,195,062, issued on Dec. 7, 2021.
Claims priority of provisional application 62/822,427, filed on Mar. 22, 2019.
Claims priority of provisional application 62/767,955, filed on Nov. 15, 2018.
Prior Publication US 2023/0260628 A1, Aug. 17, 2023
Int. Cl. G06K 9/00 (2022.01); G06F 18/2431 (2023.01); G06N 3/04 (2023.01); G06T 7/00 (2017.01); G06T 7/10 (2017.01); G06V 20/69 (2022.01); G16H 30/20 (2018.01); G16H 30/40 (2018.01); G16H 50/20 (2018.01)
CPC G16H 30/40 (2018.01) [G06F 18/2431 (2023.01); G06N 3/04 (2013.01); G06T 7/0012 (2013.01); G06T 7/10 (2017.01); G06V 20/698 (2022.01); G16H 30/20 (2018.01); G16H 50/20 (2018.01); G06T 2207/20084 (2013.01); G06T 2207/30024 (2013.01); G06T 2207/30096 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A computer-based system for distinguishing types of cancer cells in digital image data, the system comprising:
at least one computer readable memory storing software instructions;
at least one processor coupled with the memory and, upon execution of the software instructions, performs the following operations:
receiving a digital image of a tissue sample;
identifying a set of regions of interest (RoI) in the digital image;
generating a set of feature cluster densities for the set of RoIs;
selecting classifiers for the set of RoIs from a stack of classifiers based on the set of feature cluster densities, wherein classifiers are indexed by feature cluster density;
generating a classified output for the set of RoIs based on the selected classifiers; and
identifying cancer cell types in the set of RoIs based on the classified output.