Machine learning of three-dimensional right ventricular motion enables outcome prediction in pulmonary hypertension: a cardiac MR imaging study

File Description SizeFormat 
radiol.2016161315.pdfPublished version1.74 MBAdobe PDFView/Open
Title: Machine learning of three-dimensional right ventricular motion enables outcome prediction in pulmonary hypertension: a cardiac MR imaging study
Authors: Dawes, T
Simoes monteiro de marvao, A
Shi, W
Fletcher, T
Watson, G
Wharton, J
Rhodes, C
Howard, L
Gibbs, J
Rueckert, D
Cook, S
Wilkins, M
O'Regan, DP
Item Type: Journal Article
Abstract: Purpose: To determine if patient survival and mechanisms of right ventricular (RV) failure in pulmonary hypertension (PH) could be predicted usin g supervised machine learning of three dimensional patterns of systolic cardiac motion. Materials and methods: The study was approved by a research ethics committ ee and participants gave written informed consent. 256 patients (143 females, mean a ge 63 ± 17) with newly diagnosed PH underwent cardiac MR imaging, right heart catheteri zation (RHC) and six minute walk testing (6MWT) with a median follow up of 4.0 years . Semi automated segmentation of short axis cine images was used to create a three dimensi onal model of right ventricular motion. Supervised principal components analysis identified patterns of systolic motion which were most strongly predictive of survival. Survival pred iction was assessed by the difference in median survival time and the area under the curve ( AUC) using time dependent receiver operator characteristic for one year survival. Results: At the end of follow up 33% (93/256) died and one u nderwent lung transplantation. Poor outcome was predicted by a loss of effective contra ction in the septum and freewall coupled with reduced basal longitudinal motion. When added to conventional imaging, hemodynamic, functional and clinical markers, three dimensional cardiac motion improved survival prediction (area under the curve 0.73 vs 0.60, p<0. 001) and provided greater differentiation by difference in median survival time between high and low risk groups (13.8 vs 10.7 years, p<0.001). Conclusion: Three dimensional motion modeling with machine lear ning approaches reveal the adaptations in function that occur early in right heart failure and independently predict outcomes in newly diagnosed PH patients.
Issue Date: 1-May-2017
Date of Acceptance: 16-Nov-2016
ISSN: 1527-1315
Publisher: Radiological Society of North America (RSNA)
Start Page: 381
End Page: 390
Journal / Book Title: Radiology
Volume: 283
Issue: 2
Sponsor/Funder: GlaxoSmithKline Services Unlimited
National Institute for Health Research
Engineering & Physical Science Research Council (EPSRC)
British Heart Foundation
Funder's Grant Number: COL011953
RDB02 79560
Keywords: Science & Technology
Life Sciences & Biomedicine
Radiology, Nuclear Medicine & Medical Imaging
Heart Ventricles
Hypertension, Pulmonary
Image Interpretation, Computer-Assisted
Imaging, Three-Dimensional
Machine Learning
Magnetic Resonance Imaging, Cine
Middle Aged
Pattern Recognition, Automated
Reproducibility of Results
Sensitivity and Specificity
Stroke Volume
Ventricular Dysfunction, Right
Nuclear Medicine & Medical Imaging
11 Medical And Health Sciences
Publication Status: Published
Appears in Collections:Faculty of Engineering
Clinical Sciences
Imaging Sciences
National Heart and Lung Institute
Molecular Sciences
Department of Medicine (up to 2019)
Faculty of Medicine

Items in Spiral are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons