Skip to main content

Table 3 Sensitivity (Recall) evaluating predictions from UNet, ColorAE, and ensemble methods against dilated eed labels

From: Deep learning-based image analysis methods for brightfield-acquired multiplex immunohistochemistry images

Model

CD3

CD4

CD8

CD20

CD16

K17

ColorAE

0.974

0.798

0.868

0.212

0.914

0.871

UNet

0.931

0.675

0.881

0.232

0.991

0.756

Intersection

0.919

0.627

0.865

0.163

0.909

0.727

Union

0.981

0.848

0.887

0.275

0.993

0.909

Union anchor AE

0.974

0.795

0.872

0.212

0.936

0.887

Union anchor UNet

0.933

0.686

0.884

0.229

0.992

0.846