Skip to main content
. 2022 Mar 28;88(1):464–475. doi: 10.1002/mrm.29215

FIGURE 1.

MRM-29215-FIG-0001-c

Schematic illustration of the multi‐contrast data used for generating the custom set of body models (N = 10) to serve as ground truth, of which the T1‐weighted data is used as input for training the deep learning method. Whereas the semi‐automatic segmentation process involves many steps with elaborate user interaction, the deep learning method produces the body model from the original T1‐weighted data automatically