US 12,347,569 B2
Systems and methods for multiple instance learning for classification and localization in biomedical imaging
Thomas Fuchs, New York, NY (US); and Gabriele Campanella, New York, NY (US)
Assigned to Memorial Sloan-Kettering Cancer Center, New York, NY (US)
Filed by Memorial Sloan-Kettering Cancer Center, New York, NY (US)
Filed on Sep. 28, 2023, as Appl. No. 18/476,613.
Application 18/476,613 is a continuation of application No. 17/985,114, filed on Nov. 10, 2022, granted, now 11,810,677.
Application 17/985,114 is a continuation of application No. 17/074,293, filed on Oct. 19, 2020, granted, now 11,538,155, issued on Dec. 27, 2022.
Application 17/074,293 is a continuation of application No. 16/599,992, filed on Oct. 11, 2019, granted, now 10,810,736, issued on Oct. 20, 2020.
Application 16/599,992 is a continuation of application No. 16/362,470, filed on Mar. 22, 2019, granted, now 10,445,879, issued on Oct. 15, 2019.
Claims priority of provisional application 62/670,432, filed on May 11, 2018.
Claims priority of provisional application 62/647,002, filed on Mar. 23, 2018.
Prior Publication US 2024/0021324 A1, Jan. 18, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 7/00 (2017.01); G06F 18/21 (2023.01); G06F 18/2113 (2023.01); G06F 18/2415 (2023.01); G06F 18/2431 (2023.01); G06N 20/00 (2019.01); G06V 10/762 (2022.01); G06V 10/764 (2022.01); G06V 10/98 (2022.01); G16H 30/40 (2018.01); G16H 50/70 (2018.01)
CPC G16H 50/70 (2018.01) [G06F 18/2113 (2023.01); G06F 18/217 (2023.01); G06F 18/2415 (2023.01); G06F 18/2431 (2023.01); G06N 20/00 (2019.01); G06T 7/0012 (2013.01); G06V 10/7635 (2022.01); G06V 10/764 (2022.01); G06V 10/98 (2022.01); G16H 30/40 (2018.01); G06T 2207/10056 (2013.01); G06T 2207/20076 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30096 (2013.01); G06V 2201/03 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for classifying biomedical images, comprising:
identifying a plurality of tiles of a biomedical image;
applying an inference machine learning model to each tile of the plurality of tiles to generate a score indicating a likelihood of one of a presence or an absence of a condition in a corresponding tile of the plurality of tiles;
selecting a subset of tiles from the plurality of tiles based on the score of each tile of the plurality of tiles, wherein the subset of tiles includes only tiles from the plurality of tiles; and
applying an aggregation machine learning model, different from the inference machine learning model, to the subset of tiles to determine a classification result by classifying the biomedical image as either having the presence of the condition or having the absence of the condition.