Figure 3 | Scientific Reports

Figure 3

From: High spatially sensitive quantitative phase imaging assisted with deep neural network for classification of human spermatozoa under stressed condition

Figure 3

Workflow diagram showing the important steps for the classification of quantitative phase map of sperm cells. Phase map of the images is reconstructed by the interferogram captured using PSC-DHM system. Classification of the phase images is done by total 7 deep neural networks (DNN). Each network is trained with total 6,720 phase images and 2,880 phase images are used to test its accuracy. The final performance and sensitivity of the network is reported in term of confusion matrix.

Back to article page