US 11,859,253 B2
Identifying candidate cells using image analysis
Huangpin B. Hsieh, Palo Alto, CA (US); XiaoMing Wang, Dublin, CA (US); Jr-Ming Lai, Taipei (TW); Rui Mei, Santa Clara, CA (US); Hung-Jen Shao, Taipei (TW); and Jen-Chia Wu, Taipei (TW)
Assigned to CellMax Ltd., Sunnyvale, CA (US)
Filed by CellMax Ltd., Grand Cayman (KY)
Filed on Jan. 13, 2020, as Appl. No. 16/740,799.
Application 16/740,799 is a continuation of application No. 15/998,990, filed on Aug. 20, 2018, granted, now 10,533,230.
Application 15/998,990 is a continuation of application No. 15/632,707, filed on Jun. 26, 2017, granted, now 10,053,739, issued on May 21, 2018.
Application 15/632,707 is a continuation of application No. 15/476,848, filed on Mar. 31, 2017, granted, now 9,738,937, issued on Aug. 22, 2017.
Prior Publication US 2020/0325541 A1, Oct. 15, 2020
This patent is subject to a terminal disclaimer.
Int. Cl. G06K 9/00 (2022.01); C12Q 1/6886 (2018.01); G01N 15/14 (2006.01); G01N 33/574 (2006.01); G01N 33/50 (2006.01); G01N 1/31 (2006.01); G01N 33/58 (2006.01); C12Q 1/6883 (2018.01); G06T 7/00 (2017.01)
CPC C12Q 1/6886 (2013.01) [G01N 1/31 (2013.01); G01N 15/1475 (2013.01); G01N 33/5094 (2013.01); G01N 33/57484 (2013.01); G01N 33/582 (2013.01); C12Q 1/6883 (2013.01); G01N 15/1468 (2013.01); G01N 2015/1472 (2013.01); G01N 2333/4742 (2013.01); G01N 2333/70589 (2013.01); G06T 7/0012 (2013.01); G06T 2207/10056 (2013.01); G06T 2207/30096 (2013.01); G06T 2207/30101 (2013.01)] 12 Claims
OG exemplary drawing
 
1. A computer program product for identifying candidate target cells within a biological fluid specimen, the computer program product tangibly embodied in a non-transitory computer readable medium, comprising instructions to cause a processor to:
receive a digital image of the biological fluid specimen;
identify at least one candidate region of pixels in the digital image having the shape of a cell, wherein the instructions to identify the at least one candidate region having the shape of a cell include instructions to identify connected regions of pixels of a minimum intensity having a size between a minimum size and a maximum size and an aspect ratio that meets a threshold; and
determine whether a single candidate region having the shape is identified or a plurality of candidate regions having the shape are identified in the image; and
if a single candidate is identified then treat the candidate region as an image artifact, whereas if a plurality of candidate regions are identified then continue to treat the portions of the image corresponding to the plurality of candidate regions as candidates for classification.