Skip to main content
. 2019 Sep 14;20:472. doi: 10.1186/s12859-019-3037-5

Fig. 11.

Fig. 11

Comparison between FRCN learning with and without auxiliary data. For each target dateset (i.e., cervix, colorectum or kidney), all its training data are mixed with different numbers of auxiliary training data (x-axis represents the percentage of auxiliary training data)