US 12,094,126 B2
Tracking biological objects over time and space
Erick K. Moen, El Segundo, CA (US); Enrico Z. Borba, Pasadena, CA (US); David A. Van Valen, Pasadena, CA (US); William H. Graf, Pasadena, CA (US); and Barry D. Bannon, Los Angeles, CA (US)
Assigned to California Institute of Technology, Pasadena, CA (US)
Filed by California Institute of Technology, Pasadena, CA (US)
Filed on Dec. 30, 2022, as Appl. No. 18/148,992.
Application 18/148,992 is a continuation of application No. 16/859,885, filed on Apr. 27, 2020, granted, now 11,544,843.
Claims priority of provisional application 62/839,513, filed on Apr. 26, 2019.
Prior Publication US 2023/0394675 A1, Dec. 7, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 7/00 (2017.01); G06T 7/194 (2017.01); G06T 7/20 (2017.01); G06T 7/90 (2017.01); G06T 11/00 (2006.01); G06V 10/764 (2022.01); G06V 10/774 (2022.01); G06V 10/776 (2022.01); G06V 10/82 (2022.01); G06V 10/94 (2022.01); G06V 20/69 (2022.01)
CPC G06T 7/194 (2017.01) [G06T 7/0012 (2013.01); G06T 7/20 (2013.01); G06T 7/90 (2017.01); G06T 11/001 (2013.01); G06V 10/764 (2022.01); G06V 10/774 (2022.01); G06V 10/776 (2022.01); G06V 10/82 (2022.01); G06V 10/945 (2022.01); G06V 20/698 (2022.01); G06T 2200/24 (2013.01); G06T 2207/10024 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30024 (2013.01)] 23 Claims
OG exemplary drawing
 
1. A system for training a deep learning model for cell tracking and lineage construction, comprising:
non-transitory memory configured to store executable instructions; and
a hardware processor in communication with the non-transitory memory, the hardware processor programmed by the executable instructions to perform:
generating training data, or receiving training data generated, from raw images, frame numbers associated with the raw images, associations, each comprising a cell mask having pixels and an associated cell label assigned to one biological object in the raw images, and a new association, wherein the new association is determined by:
generating, or causing display, or both, a user interface comprising (i) a first panel for displaying a frame of at least one of (ia) at least one raw image of the raw images at a time, (ib) a mask image comprising one or more cell masks associated with one or more biological objects in the at least one raw image, and (ic) an outline corresponding to boundaries of the one or more cell masks associated with the one or more biological objects in the at least one raw image, and (ii) a second panel for displaying at least one of (iia) the frame number of the at least one raw image being displayed, (iib) tracking information associated with the one or more biological objects, and (iic) cell selection information of one or more cells selected;
receiving a selection of one or more new cell mask pixels of the at least one raw image being displayed;
generating, or causing display, or both, an updated user interface comprising (i) the first panel displaying the frame of the at least one raw image of the raw images and the one or more new cell mask pixels highlighted; and
generating the new association of (iiia) the new cell mask having the one or more new cell mask pixels of the at least one raw image, and (iiib) a new cell label; and
training a deep learning model for cell tracking and lineage construction using the training data generated,
wherein the system comprises pods, or wherein the system dynamically allocates the computational resources using horizontal pod scaling, or both.