Skip to main content
. Author manuscript; available in PMC: 2022 Nov 1.
Published in final edited form as: J Cardiovasc Comput Tomogr. 2021 Mar 22;15(6):462–469. doi: 10.1016/j.jcct.2021.03.006

Figure 1. Automatic segmentation of cardiac structures using deep learning.

Figure 1.

Cardiac CT acquisitions from a dual-layer detector CT scanner were reconstructed into coronary CT angiography (CCTA), corresponding virtual noncontrast (VNC), and true noncontrast CT (NCCT) images. Manual reference segmentations of 7 cardiac structures on CCTA images (left) were used to train CNNs for automatic segmentation in VNC (middle) and NCCT (right) images. Shown is a case example of DL-based segmentations in axial (top), sagittal (middle), and coronal (bottom) views. Reproduced with permission from Bruns et al.15