3D Pulmonary Artery Segmentation from CTA Scans Using Deep Learning with Realistic Data Augmentation
Authors: Isaac de la Bruere Jorge Onieva Lasse Andresen Jakob Qvortrup Holsting Farbod Nicholas Rahaghi Miguel Ángel González Ballester Raúl San José Estepar
Date: 01.09.2018
Abstract
The characterization of the vasculature in the mediastinum, more specifically the pulmonary artery, is of vital importance for the evaluation of several pulmonary vascular diseases. Thus, the goal of this study is to automatically segment the pulmonary artery (PA) from computed tomography angiography images, which opens up the opportunity for more complex analysis of the evolution of the PA geometry in health and disease and can be used in complex fluid mechanics models or individualized medicine. For that purpose, a new 3D convolutional neural network architecture is proposed, which is trained on images coming from different patient cohorts. The network makes use a strong data augmentation paradigm based on realistic deformations generated by applying principal component analysis to the deformation fields obtained from the affine registration of several datasets. The network is validated on 91 datasets by comparing the automatic segmentations with semi-automatically delineated ground truths in terms of mean Dice and Jaccard coefficients and mean distance between surfaces, which yields values of 0.89, 0.80 and 1.25 mm, respectively. Finally, a comparison against a Unet architecture is also included.
BIB_text
title = {3D Pulmonary Artery Segmentation from CTA Scans Using Deep Learning with Realistic Data Augmentation},
pages = {225-237},
keywds = {
Pulmonary artery, Deep learning, CTA, Convolutional neural network, Segmentation
}
abstract = {
The characterization of the vasculature in the mediastinum, more specifically the pulmonary artery, is of vital importance for the evaluation of several pulmonary vascular diseases. Thus, the goal of this study is to automatically segment the pulmonary artery (PA) from computed tomography angiography images, which opens up the opportunity for more complex analysis of the evolution of the PA geometry in health and disease and can be used in complex fluid mechanics models or individualized medicine. For that purpose, a new 3D convolutional neural network architecture is proposed, which is trained on images coming from different patient cohorts. The network makes use a strong data augmentation paradigm based on realistic deformations generated by applying principal component analysis to the deformation fields obtained from the affine registration of several datasets. The network is validated on 91 datasets by comparing the automatic segmentations with semi-automatically delineated ground truths in terms of mean Dice and Jaccard coefficients and mean distance between surfaces, which yields values of 0.89, 0.80 and 1.25 mm, respectively. Finally, a comparison against a Unet architecture is also included.
}
isbn = {978-303000945-8},
doi = {10.1007/978-3-030-00946-5_23},
date = {2018-09-01},
}
