Skip to main content

Table 2 F1 Scores evaluating predictions from UNet, ColorAE, and ensemble methods against dilated seed labels

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

Model CD3 CD4 CD8 CD20 CD16 K17
ColorAE 0.652 0.728 0.72 0.347 0.796 0.637
UNet 0.628 0.661 0.628 0.353 0.687 0.739
Intersection 0.698 0.693 0.722 0.279 0.79 0.749
Union 0.593 0.705 0.628 0.403 0.691 0.638
Union anchor AE 0.662 0.732 0.731 0.346 0.836 0.657
Union anchor UNet 0.639 0.674 0.633 0.35 0.7 0.806