TY - JOUR
T1 - Deep learning-based whole-heart segmentation in 4D contrast-enhanced cardiac CT
AU - Bruns, Steffen
AU - Wolterink, Jelmer M.
AU - van den Boogert, Thomas P.W.
AU - Runge, Jurgen H.
AU - Bouma, Berto J.
AU - Henriques, José P.
AU - Baan, Jan
AU - Viergever, Max A.
AU - Planken, R. Nils
AU - Išgum, Ivana
N1 - Funding Information: Ivana Išgum received institutional research projects by the Dutch Technology Foundation co-funded by Pie Medical Imaging and Philips Healthcare ( P15-26 ), the Netherlands Organisation for Health Research and Development with participation of Pie Medical Imaging ( 104003009 ), institutional research grant by Dutch Cancer Foundation (BRAGATSTON study), institutional research grant by Dutch Heart Foundation ( CVON2015-17 ), and institutional research grants funded by Pie Medical Imaging. Ivana Išgum is a cofounder and scientific lead of Quantib BV, the Netherlands. Jelmer M. Wolterink acknowledges funding from the 4TU Precision Medicine program supported by High Tech for a Sustainable Future, a framework commissioned by the four Universities of Technology of the Netherlands. Jan Baan received an independent research grant from Edwards Lifesciences . Berto J. Bouma received an unrestricted research grant from Abbott . The authors declare that the research was conducted in the absence of any additional commercial or financial relationships that could be construed as a potential conflict of interest. Funding Information: This study was funded by the Dutch Technology Foundation ( STW , perspectief, P15-26) with participation of Philips Healthcare, Haifa, Israel. Publisher Copyright: © 2021 The Authors
PY - 2022/3/1
Y1 - 2022/3/1
N2 - Automatic cardiac chamber and left ventricular (LV) myocardium segmentation over the cardiac cycle significantly extends the utilization of contrast-enhanced cardiac CT, potentially enabling in-depth assessment of cardiac function. Therefore, we evaluate an automatic method for cardiac chamber and LV myocardium segmentation in 4D cardiac CT. In this study, 4D contrast-enhanced cardiac CT scans of 1509 patients selected for transcatheter aortic valve implantation with 21,605 3D images, were divided into development (N = 12) and test set (N = 1497). 3D convolutional neural networks were trained with end-systolic (ES) and end-diastolic (ED) images. Dice similarity coefficient (DSC) and average symmetric surface distance (ASSD) were computed for 3D segmentations at ES and ED in the development set via cross-validation, and for 2D segmentations in four cardiac phases for 81 test set patients. Segmentation quality in the full test set of 1497 patients was assessed visually on a three-point scale per structure based on estimated overlap with the ground truth. Automatic segmentation resulted in a mean DSC of 0.89 ± 0.10 and ASSD of 1.43 ± 1.45 mm in 12 patients in 3D, and a DSC of 0.89 ± 0.08 and ASSD of 1.86 ± 1.20 mm in 81 patients in 2D. The qualitative evaluation in the whole test set of 1497 patients showed that automatic segmentations were assigned grade 1 (clinically useful) in 98.5%, 92.2%, 83.1%, 96.3%, and 91.6% of cases for LV cavity and myocardium, right ventricle, left atrium, and right atrium. Our automatic method using convolutional neural networks performed clinically useful segmentation across the cardiac cycle in a large set of 4D cardiac CT images, potentially enabling in-depth assessment of cardiac function.
AB - Automatic cardiac chamber and left ventricular (LV) myocardium segmentation over the cardiac cycle significantly extends the utilization of contrast-enhanced cardiac CT, potentially enabling in-depth assessment of cardiac function. Therefore, we evaluate an automatic method for cardiac chamber and LV myocardium segmentation in 4D cardiac CT. In this study, 4D contrast-enhanced cardiac CT scans of 1509 patients selected for transcatheter aortic valve implantation with 21,605 3D images, were divided into development (N = 12) and test set (N = 1497). 3D convolutional neural networks were trained with end-systolic (ES) and end-diastolic (ED) images. Dice similarity coefficient (DSC) and average symmetric surface distance (ASSD) were computed for 3D segmentations at ES and ED in the development set via cross-validation, and for 2D segmentations in four cardiac phases for 81 test set patients. Segmentation quality in the full test set of 1497 patients was assessed visually on a three-point scale per structure based on estimated overlap with the ground truth. Automatic segmentation resulted in a mean DSC of 0.89 ± 0.10 and ASSD of 1.43 ± 1.45 mm in 12 patients in 3D, and a DSC of 0.89 ± 0.08 and ASSD of 1.86 ± 1.20 mm in 81 patients in 2D. The qualitative evaluation in the whole test set of 1497 patients showed that automatic segmentations were assigned grade 1 (clinically useful) in 98.5%, 92.2%, 83.1%, 96.3%, and 91.6% of cases for LV cavity and myocardium, right ventricle, left atrium, and right atrium. Our automatic method using convolutional neural networks performed clinically useful segmentation across the cardiac cycle in a large set of 4D cardiac CT images, potentially enabling in-depth assessment of cardiac function.
KW - 4D contrast-enhanced cardiac CT
KW - Convolutional neural network
KW - Deep learning
KW - Transcatheter aortic valve implantation
KW - Whole-heart segmentation
UR - http://www.scopus.com/inward/record.url?scp=85122527310&partnerID=8YFLogxK
UR - https://pure.uva.nl/ws/files/67542886/1_s2.0_S0010482521009859_mmc1.mp4
U2 - https://doi.org/10.1016/j.compbiomed.2021.105191
DO - https://doi.org/10.1016/j.compbiomed.2021.105191
M3 - Article
C2 - 35026571
SN - 0010-4825
VL - 142
JO - Computers in Biology and Medicine
JF - Computers in Biology and Medicine
M1 - 105191
ER -