Table 3.

Effect of preprocessing on metrics of the training dataset

Condition	Precision	Recall	Dice
S0033	0.822–0.814 (0.008)	0.885–0.879 (0.006)	0.853–0.846 (0.007)
S0066	0.868–0.748 (0.120)	0.896–0.839 (0.057)	0.882–0.791 (0.091)
S0150	0.883–0.839 (0.044)	0.896–0.881 (0.015)	0.889–0.858 (0.031)
S0300	0.881–0.861 (0.020)	0.905–0.908 (− 0.003)	0.893–0.884 (0.009)
S0600	0.917–0.788 (0.129)	0.888–0.872 (0.061)	0.902–0.828 (0.074)
S3600	0.895–0.892 (0.003)	0.924–0.904 (0.020)	0.909–0.898 (0.011)

Assuming that the metric value with processing is A and the value without processing is B, the table is presented as A—B (difference)