Taeouk Kim1, Mohammadali Hedayat1, Veronica V Vaitkus2, Marek Belohlavek2, Vinayak Krishnamurthy1,3, Iman Borazjani1. 1. J. Mike Walker '66, Department of Mechanical Engineering, Texas A&M University, College Station, Texas, USA. 2. Department of Cardiovascular Diseases, Mayo Clinic, Scottsdale, Arizona, USA. 3. Department of Computer Science (By Affiliation), Texas A&M University, College Station, Texas, USA.
Abstract
BACKGROUND: Two-dimensional echocardiography (2D echo) is the most widely used non-invasive imaging modality due to its fast acquisition time, low cost, and high temporal resolution. Boundary identification of left ventricle (LV) in 2D echo, i.e., image segmentation, is the first step to calculate relevant clinical parameters. Currently, LV segmentation in 2D echo is primarily conducted semi-manually. A fully-automatic segmentation of the LV wall needs further development. METHODS: We evaluated the performance of the state-of-the-art convolutional neural networks (CNNs) for the segmentation of 2D echo images from 6 standard projections of the LV. We used two segmentation algorithms: U-net and segAN. The models were trained using an in-house dataset, which consists of 1,649 porcine images from 6 to 8 different pigs. In addition, a transfer learning approach was used for the segmentation of long-axis projections by training models with our database based on the previously trained weights obtained from Cardiac Acquisitions for Multi-structure Ultrasound Segmentation (CAMUS) dataset. The models were tested on a separate set of images from two other pigs by computing several metrics. The segmentation process was combined with a 3D reconstruction framework to quantify the physiological indices such as LV volumes and ejection fraction (EF). RESULTS: The average dice metric for the LV cavity was 0.90 and 0.91 for the U-net and segAN, respectively, which was higher than 0.82 for the level-set (P value: 3.31×10-25). The average Hausdorff distance for the LV cavity was 2.71 mm and 2.82 mm for the U-net and segAN, respectively, which was lower than 3.64 mm for the level-set (P value: 4.86×10-16). The LV shapes and volumes obtained using the CNN segmentation models were in good agreement with the results segmented by the experts. In addition, the differences of the calculated physiological parameters between two 3D reconstruction models segmented by the experts and CNNs were less than 15%. CONCLUSIONS: The results showed that both CNN models achieve higher performance on LV segmentation than the level-set method. The error of the reconstruction from automatic segmentation compared to the expert segmentation is less than 15%, which is within the 20% error of echo compared to the gold standard. 2021 Quantitative Imaging in Medicine and Surgery. All rights reserved.
BACKGROUND: Two-dimensional echocardiography (2D echo) is the most widely used non-invasive imaging modality due to its fast acquisition time, low cost, and high temporal resolution. Boundary identification of left ventricle (LV) in 2D echo, i.e., image segmentation, is the first step to calculate relevant clinical parameters. Currently, LV segmentation in 2D echo is primarily conducted semi-manually. A fully-automatic segmentation of the LV wall needs further development. METHODS: We evaluated the performance of the state-of-the-art convolutional neural networks (CNNs) for the segmentation of 2D echo images from 6 standard projections of the LV. We used two segmentation algorithms: U-net and segAN. The models were trained using an in-house dataset, which consists of 1,649 porcine images from 6 to 8 different pigs. In addition, a transfer learning approach was used for the segmentation of long-axis projections by training models with our database based on the previously trained weights obtained from Cardiac Acquisitions for Multi-structure Ultrasound Segmentation (CAMUS) dataset. The models were tested on a separate set of images from two other pigs by computing several metrics. The segmentation process was combined with a 3D reconstruction framework to quantify the physiological indices such as LV volumes and ejection fraction (EF). RESULTS: The average dice metric for the LV cavity was 0.90 and 0.91 for the U-net and segAN, respectively, which was higher than 0.82 for the level-set (P value: 3.31×10-25). The average Hausdorff distance for the LV cavity was 2.71 mm and 2.82 mm for the U-net and segAN, respectively, which was lower than 3.64 mm for the level-set (P value: 4.86×10-16). The LV shapes and volumes obtained using the CNN segmentation models were in good agreement with the results segmented by the experts. In addition, the differences of the calculated physiological parameters between two 3D reconstruction models segmented by the experts and CNNs were less than 15%. CONCLUSIONS: The results showed that both CNN models achieve higher performance on LV segmentation than the level-set method. The error of the reconstruction from automatic segmentation compared to the expert segmentation is less than 15%, which is within the 20% error of echo compared to the gold standard. 2021 Quantitative Imaging in Medicine and Surgery. All rights reserved.
Entities:
Keywords:
3D reconstruction; Convolutional neural networks (CNNs); deep learning; echocardiography; left ventricle (LV)
Authors: Roberto M Lang; Luigi P Badano; Wendy Tsang; David H Adams; Eustachio Agricola; Thomas Buck; Francesco F Faletra; Andreas Franke; Judy Hung; Leopoldo Pérez de Isla; Otto Kamp; Jaroslaw D Kasprzak; Patrizio Lancellotti; Thomas H Marwick; Marti L McCulloch; Mark J Monaghan; Petros Nihoyannopoulos; Natesa G Pandian; Patricia A Pellikka; Mauro Pepi; David A Roberson; Stanton K Shernan; Girish S Shirali; Lissa Sugeng; Folkert J Ten Cate; Mani A Vannan; Jose Luis Zamorano; William A Zoghbi Journal: Eur Heart J Cardiovasc Imaging Date: 2012-01 Impact factor: 6.875
Authors: Chris A Cocosco; Wiro J Niessen; Thomas Netsch; Evert-Jan P A Vonken; Gunnar Lund; Alexander Stork; Max A Viergever Journal: J Magn Reson Imaging Date: 2008-08 Impact factor: 4.813
Authors: Johannes Greupner; Elke Zimmermann; Andrea Grohmann; Hans-Peter Dübel; Till F Althoff; Till Althoff; Adrian C Borges; Wolfgang Rutsch; Peter Schlattmann; Bernd Hamm; Marc Dewey Journal: J Am Coll Cardiol Date: 2012-05-22 Impact factor: 24.094
Authors: Mohammadali Hedayat; Tatsat R Patel; Taeouk Kim; Marek Belohlavek; Kenneth R Hoffmann; Iman Borazjani Journal: Int J Numer Method Biomed Eng Date: 2020-06-08 Impact factor: 2.747
Authors: Avan Suinesiaputra; Brett R Cowan; Ahmed O Al-Agamy; Mustafa A Elattar; Nicholas Ayache; Ahmed S Fahmy; Ayman M Khalifa; Pau Medrano-Gracia; Marie-Pierre Jolly; Alan H Kadish; Daniel C Lee; Ján Margeta; Simon K Warfield; Alistair A Young Journal: Med Image Anal Date: 2013-09-13 Impact factor: 8.545
Authors: Honghai Zhang; Andreas Wahle; Ryan K Johnson; Thomas D Scholz; Milan Sonka Journal: IEEE Trans Med Imaging Date: 2009-08-25 Impact factor: 10.048
Authors: Ursula Reiter; Gert Reiter; Martin Manninger; Gabriel Adelsmayr; Julia Schipke; Alessio Alogna; Alexandra Rajces; Aurelien F Stalder; Andreas Greiser; Christian Mühlfeld; Daniel Scherr; Heiner Post; Burkert Pieske; Michael Fuchsjäger Journal: J Cardiovasc Magn Reson Date: 2016-09-30 Impact factor: 5.364