Skip to main content
. 2022 Jan 11;1:796078. doi: 10.3389/fradi.2021.796078

Table 6.

Comparison of the multi-task training to the mono-task on the DHMItest dataset; “-” stands for unavailable metrics.

Task
breastbin ACR BIRADS View
AUC Accuracy Accuracy AUC
MTL 80.46 ± 0.79 55.95 ± 1.12 67.66 ± 1.06 96.55 ± 0.70
breast bin 70.69 ± 0.36 - - -
ACR - 54.67 ± 1.13 - -
BIRADS - - 65.92 ± 0.96 -
view - - - 97.14 ± 0.46