US 12,266,104 B2
Fluorescence lifetime imaging using deep learning
Jason Tyler Smith, Troy, NY (US); Ruoyang Yao, Elmhurst, NY (US); Xavier Intes, Schenectady, NY (US); Pingkun Yan, Clifton Park, NY (US); and Marien Ochoa-Mendoza, Troy, NY (US)
Assigned to Rensselaer Polytechnic Institute, Troy, NY (US)
Filed by Rensselaer Polytechnic Institute, Troy, NY (US)
Filed on Jan. 26, 2024, as Appl. No. 18/423,472.
Application 18/423,472 is a continuation of application No. 17/143,448, filed on Jan. 7, 2021, granted, now 11,887,298.
Claims priority of provisional application 63/134,536, filed on Jan. 6, 2021.
Claims priority of provisional application 63/001,947, filed on Mar. 30, 2020.
Claims priority of provisional application 62/958,022, filed on Jan. 7, 2020.
Prior Publication US 2024/0193768 A1, Jun. 13, 2024
Int. Cl. G06T 7/00 (2017.01); G06N 3/045 (2023.01); G06N 3/088 (2023.01)
CPC G06T 7/0012 (2013.01) [G06N 3/045 (2023.01); G06N 3/088 (2013.01); G06T 2207/10064 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30004 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus for fluorescence lifetime imaging (FLI), the apparatus comprising:
a deep neural network (DNN) comprising:
a first convolutional layer configured to receive FLI input data;
a plurality of intermediate layers, each intermediate layer configured to receive a respective intermediate input corresponding to an output of a respective prior layer, each intermediate layer further configured to provide a respective intermediate output related to the received respective intermediate input; and
an output layer configured to provide estimated FLI output data corresponding to the received FLI input data;
wherein the DNN is trained using synthetic data; and
a discriminator network configured to compare estimated training output data with training synthetic output data during training, the DNN and the discriminator network corresponding to a generative adversarial network (GAN) during training, the discriminator network comprising:
a first two-dimensional (2D) convolutional block;
a first intermediate block coupled to an output of the first 2D convolutional block;
a second 2D convolutional block coupled to an output of the first intermediate block;
a second intermediate block coupled to an output of the second 2D convolutional block;
a flatten block coupled to an output of the second intermediate block;
a first dense block coupled to an output of the flatten block;
a third intermediate block coupled to an output of the first dense block;
a second dense block coupled to an output of the third intermediate block;
a dropout block coupled to an output of the second dense block; and
a sigmoid block coupled to an output of the dropout block;
wherein the first intermediate block, the second intermediate block, and the third intermediate block each comprise a batch normalization function and a rectified linear unit.