US 12,085,568 B2
Machine learning for digital pathology
David B. Agus, Beverly Hills, CA (US); Paul Thomas Macklin, Bloomington, IN (US); Rishi Raghav Rawat, Boulder, CO (US); and Daniel Lee Ruderman, Los Angeles, CA (US)
Assigned to University of Southern California, Los Angeles, CA (US)
Filed by University of Southern California, Los Angeles, CA (US)
Filed on Nov. 7, 2023, as Appl. No. 18/387,608.
Application 18/387,608 is a continuation of application No. 16/491,700, granted, now 11,835,524, previously published as PCT/US2018/021060, filed on Mar. 6, 2018.
Claims priority of provisional application 62/467,579, filed on Mar. 6, 2017.
Prior Publication US 2024/0069026 A1, Feb. 29, 2024
Int. Cl. G06K 9/00 (2022.01); G01N 33/574 (2006.01); G06F 18/211 (2023.01); G06T 7/00 (2017.01); G06V 10/82 (2022.01); G06V 20/69 (2022.01); G16H 50/20 (2018.01)
CPC G01N 33/574 (2013.01) [G06F 18/211 (2023.01); G06T 7/0012 (2013.01); G06V 10/82 (2022.01); G06V 20/695 (2022.01); G06V 20/698 (2022.01); G16H 50/20 (2018.01); G06T 2207/10024 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30024 (2013.01)] 22 Claims
OG exemplary drawing
 
1. A method for determining a status of a diagnostic, prognostic, or theragnostic feature of a test stained tissue sample, comprising:
(a) obtaining a sample digital image for the test stained tissue sample, the sample digital image comprising an associated set of spatial locations;
(b) processing the sample digital image to generate a set of extracted features therefrom, wherein the set of extracted features comprises a morphometric feature of the test stained tissue sample;
(c) associating a value for each extracted feature of the set of extracted features with each spatial location of the sample digital image to form a test set of extracted feature maps for the test stained tissue sample; and
(d) processing the test set of extracted feature maps by a trained machine learning algorithm to determine the status of the diagnostic, prognostic, or theragnostic feature for the test stained tissue sample,
wherein (d) further comprises a pre-processing operation comprising:
(i) processing the test set of extracted feature maps by a first classifier to generate a first classification value, wherein the first classifier is trained on first training data comprising input nuclear morphometric features and associated diagnostic, prognostic, or theragnostic statuses,
(ii) processing the test set of extracted feature maps by a second classifier to generate a second classification value, wherein the second classifier is trained on second training data comprising input cytoplasmic morphometric features and/or extracellular morphometric features and associated diagnostic, prognostic, or theragnostic statuses, and
(iii) comparing the first classification value and the second classification value to determine which morphometric features are more predictive of the determined status.