1
|
Park J, Kweon J, Kim YI, Back I, Chae J, Roh JH, Kang DY, Lee PH, Ahn JM, Kang SJ, Park DW, Lee SW, Lee CW, Park SW, Park SJ, Kim YH. Selective ensemble methods for deep learning segmentation of major vessels in invasive coronary angiography. Med Phys 2023; 50:7822-7839. [PMID: 37310802 DOI: 10.1002/mp.16554] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/10/2022] [Revised: 03/29/2023] [Accepted: 05/26/2023] [Indexed: 06/15/2023] Open
Abstract
BACKGROUND Invasive coronary angiography (ICA) is a primary imaging modality that visualizes the lumen area of coronary arteries for diagnosis and interventional guidance. In the current practice of quantitative coronary analysis (QCA), semi-automatic segmentation tools require labor-intensive and time-consuming manual correction, limiting their application in the catheterization room. PURPOSE This study aims to propose rank-based selective ensemble methods that improve the segmentation performance and reduce morphological errors that limit fully automated quantification of coronary artery using deep-learning segmentation of ICA. METHODS Two selective ensemble methods proposed in this work integrated the weighted ensemble approach with per-image quality estimation. The segmentation outcomes from five base models with different loss functions were ranked either by mask morphology or estimated dice similarity coefficient (DSC). The final output was determined by imposing different weights according to the ranks. The ranking criteria based on mask morphology were formulated from empirical insight to avoid frequent types of segmentation errors (MSEN), while the estimation of DSCs was performed by comparing the pseudo-ground truth generated from a meta-learner (ESEN). Five-fold cross-validation was performed with the internal dataset of 7426 coronary angiograms from 2924 patients, and prediction model was externally validated with 556 images of 226 patients. RESULTS The selective ensemble methods improved the segmentation performance with DSCs up to 93.07% and provided a better delineation of coronary lesion with local DSCs of up to 93.93%, outperforming all individual models. Proposed methods also minimized the chances of mask disconnection in the most narrowed regions to 2.10%. The robustness of the proposed methods was also evident in the external validation. Inference time for major vessel segmentation was approximately one-sixth of a second. CONCLUSION Proposed methods successfully reduced morphological errors in the predicted masks and were able to enhance the robustness of the automatic segmentation. The results suggest better applicability of real-time QCA-based diagnostic methods in routine clinical settings.
Collapse
Affiliation(s)
- Jeeone Park
- Department of Medical Science, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
| | - Jihoon Kweon
- Department of Convergence Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
| | - Young In Kim
- Department of Medical Science, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
| | - Inwook Back
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Jihye Chae
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Jae-Hyung Roh
- Department of Cardiology, Chungnam National University Sejong Hospital, Chungnam National University School of Medicine, Daejeon, South Korea
| | - Do-Yoon Kang
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Pil Hyung Lee
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Jung-Min Ahn
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Soo-Jin Kang
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Duk-Woo Park
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Seung-Whan Lee
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Cheol Whan Lee
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Seong-Wook Park
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Seung-Jung Park
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| | - Young-Hak Kim
- Division of Cardiology, Department of Internal Medicine, Medical Center, University of Ulsan College of Medicine, Asan, Seoul, South Korea
| |
Collapse
|
2
|
Lee YJ, Kim YW, Ha J, Kim M, Guagliumi G, Granada JF, Lee SG, Lee JJ, Cho YK, Yoon HJ, Lee JH, Kim U, Jang JY, Oh SJ, Lee SJ, Hong SJ, Ahn CM, Kim BK, Chang HJ, Ko YG, Choi D, Hong MK, Jang Y, Lee JS, Kim JS. Computational Fractional Flow Reserve From Coronary Computed Tomography Angiography—Optical Coherence Tomography Fusion Images in Assessing Functionally Significant Coronary Stenosis. Front Cardiovasc Med 2022; 9:925414. [PMID: 35770218 PMCID: PMC9234158 DOI: 10.3389/fcvm.2022.925414] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/21/2022] [Accepted: 05/25/2022] [Indexed: 11/13/2022] Open
Abstract
Background Coronary computed tomography angiography (CTA) and optical coherence tomography (OCT) provide additional functional information beyond the anatomy by applying computational fluid dynamics (CFD). This study sought to evaluate a novel approach for estimating computational fractional flow reserve (FFR) from coronary CTA-OCT fusion images. Methods Among patients who underwent coronary CTA, 148 patients who underwent both pressure wire-based FFR measurement and OCT during angiography to evaluate intermediate stenosis in the left anterior descending artery were included from the prospective registry. Coronary CTA-OCT fusion images were created, and CFD was applied to estimate computational FFR. Based on pressure wire-based FFR as a reference, the diagnostic performance of Fusion-FFR was compared with that of CT-FFR and OCT-FFR. Results Fusion-FFR was strongly correlated with FFR (r = 0.836, P < 0.001). Correlation between FFR and Fusion-FFR was stronger than that between FFR and CT-FFR (r = 0.682, P < 0.001; z statistic, 5.42, P < 0.001) and between FFR and OCT-FFR (r = 0.705, P < 0.001; z statistic, 4.38, P < 0.001). Area under the receiver operating characteristics curve to assess functionally significant stenosis was higher for Fusion-FFR than for CT-FFR (0.90 vs. 0.83, P = 0.024) and OCT-FFR (0.90 vs. 0.83, P = 0.043). Fusion-FFR exhibited 84.5% accuracy, 84.6% sensitivity, 84.3% specificity, 80.9% positive predictive value, and 87.5% negative predictive value. Especially accuracy, specificity, and positive predictive value were superior for Fusion-FFR than for CT-FFR (73.0%, P = 0.007; 61.4%, P < 0.001; 64.0%, P < 0.001) and OCT-FFR (75.7%, P = 0.021; 73.5%, P = 0.020; 69.9%, P = 0.012). Conclusion CFD-based computational FFR from coronary CTA-OCT fusion images provided more accurate functional information than coronary CTA or OCT alone. Clinical Trial Registration [www.ClinicalTrials.gov], identifier [NCT03298282].
Collapse
Affiliation(s)
- Yong-Joon Lee
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Young Woo Kim
- Department of Mechanical Engineering, Yonsei University, Seoul, South Korea
| | - Jinyong Ha
- Department of Electrical Engineering, Sejong University, Seoul, South Korea
| | - Minug Kim
- Department of Electrical Engineering, Sejong University, Seoul, South Korea
| | - Giulio Guagliumi
- Department of Cardiovascular, Ospedale Papa Giovanni XXIII, Bergamo, Italy
| | - Juan F. Granada
- Cardiovascular Research Foundation, Columbia University Medical Center, New York, NY, United States
| | - Seul-Gee Lee
- Yonsei Cardiovascular Research Institute, Yonsei University College of Medicine, Seoul, South Korea
| | - Jung-Jae Lee
- Yonsei Cardiovascular Research Institute, Yonsei University College of Medicine, Seoul, South Korea
| | - Yun-Kyeong Cho
- Department of Cardiology, Keimyung University Dongsan Hospital, Daegu, South Korea
| | - Hyuck Jun Yoon
- Department of Cardiology, Keimyung University Dongsan Hospital, Daegu, South Korea
| | - Jung Hee Lee
- Division of Cardiology, Yeungnam University Medical Center, Yeungnam University College of Medicine, Daegu, South Korea
| | - Ung Kim
- Division of Cardiology, Yeungnam University Medical Center, Yeungnam University College of Medicine, Daegu, South Korea
| | - Ji-Yong Jang
- National Health Insurance Service Ilsan Hospital, Goyang, South Korea
| | - Seung-Jin Oh
- National Health Insurance Service Ilsan Hospital, Goyang, South Korea
| | - Seung-Jun Lee
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Sung-Jin Hong
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Chul-Min Ahn
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Byeong-Keuk Kim
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Hyuk-Jae Chang
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Young-Guk Ko
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Donghoon Choi
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Myeong-Ki Hong
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
| | - Yangsoo Jang
- Division of Cardiology, CHA Bundang Medical Center, CHA University College of Medicine, Seongnam, South Korea
| | - Joon Sang Lee
- Department of Mechanical Engineering, Yonsei University, Seoul, South Korea
- *Correspondence: Joon Sang Lee,
| | - Jung-Sun Kim
- Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Seoul, South Korea
- Jung-Sun Kim,
| |
Collapse
|
3
|
Zhang J, Yang D, Lv W, Jin X, Shi Y. Three-dimensional phase and intensity reconstruction from coherent modulation imaging measurements. OPTICS EXPRESS 2022; 30:20415-20430. [PMID: 36224787 DOI: 10.1364/oe.460648] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Received: 04/07/2022] [Accepted: 05/11/2022] [Indexed: 06/16/2023]
Abstract
Coherent modulation imaging is a lensless imaging technique, where a complex-valued image can be recovered from a single diffraction pattern using the iterative algorithm. Although mostly applied in two dimensions, it can be tomographically combined to produce three-dimensional (3D) images. Here we present a 3D reconstruction procedure for the sample's phase and intensity from coherent modulation imaging measurements. Pre-processing methods to remove illumination probe, inherent ambiguities in phase reconstruction results, and intensity fluctuation are given. With the projections extracted by our method, standard tomographic reconstruction frameworks can be used to recover accurate quantitative 3D phase and intensity images. Numerical simulations and optical experiments validate our method.
Collapse
|
4
|
Ngo MT, Lee UY, Ha H, Jung J, Lee DH, Kwak HS. Improving Blood Flow Visualization of Recirculation Regions at Carotid Bulb in 4D Flow MRI Using Semi-Automatic Segmentation with ITK-SNAP. Diagnostics (Basel) 2021; 11:diagnostics11101890. [PMID: 34679588 PMCID: PMC8534781 DOI: 10.3390/diagnostics11101890] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/16/2021] [Revised: 10/06/2021] [Accepted: 10/10/2021] [Indexed: 11/16/2022] Open
Abstract
Assessment of carotid bulb hemodynamics using four-dimensional (4D) flow magnetic resonance imaging (MRI) requires accurate segmentation of recirculation regions that is frequently hampered by limited resolution. This study aims to improve the accuracy of 4D flow MRI carotid bulb segmentation and subsequent recirculation regions analysis. Time-of-flight (TOF) MRI and 4D flow MRI were performed on bilateral carotid artery bifurcations in seven healthy volunteers. TOF-MRI data was segmented into 3D geometry for computational fluid dynamics (CFD) simulations. ITK-SNAP segmentation software was included in the workflow for the semi-automatic generation of 4D flow MRI angiographic data. This study compared the velocities calculated at the carotid bifurcations and the 3D blood flow visualization at the carotid bulbs obtained by 4D flow MRI and CFD. By applying ITK-SNAP segmentation software, an obvious improvement in the 4D flow MRI visualization of the recirculation regions was observed. The 4D flow MRI images of the recirculation flow characteristics of the carotid artery bulbs coincided with the CFD. A reasonable agreement was found in terms of velocity calculated at the carotid bifurcation between CFD and 4D flow MRI. However, the dispersion of velocity data points relative to the local errors of measurement in 4D flow MRI remains. Our proposed strategy showed the feasibility of improving recirculation regions segmentation and the potential for reliable blood flow visualization in 4D flow MRI. However, quantitative analysis of recirculation regions in 4D flow MRI with ITK-SNAP should be enhanced for use in clinical situations.
Collapse
Affiliation(s)
- Minh Tri Ngo
- Department of Radiology of Hue Central Hospital, Hue, Thua Thien Hue 530000, Vietnam;
| | - Ui Yun Lee
- Division of Mechanical Design Engineering, College of Engineering, Jeonbuk National University, Jeon-ju 54896, Korea; (U.Y.L.); (J.J.)
| | - Hojin Ha
- Department of Mechanical and Biomedical Engineering, Kangwon National University, Chuncheon 24341, Korea;
| | - Jinmu Jung
- Division of Mechanical Design Engineering, College of Engineering, Jeonbuk National University, Jeon-ju 54896, Korea; (U.Y.L.); (J.J.)
- Hemorheology Research Institute, Jeonbuk National University, Jeon-ju 54896, Korea
| | - Dong Hwan Lee
- Division of Mechanical Design Engineering, College of Engineering, Jeonbuk National University, Jeon-ju 54896, Korea; (U.Y.L.); (J.J.)
- Hemorheology Research Institute, Jeonbuk National University, Jeon-ju 54896, Korea
- Correspondence: (D.H.L.); (H.S.K.); Tel.: +82-63-270-3998 (D.H.L.); +82-63-250-2582 (H.S.K.)
| | - Hyo Sung Kwak
- Department of Radiology and Research Institute of Clinical Medicine of Jeonbuk National University, Biomedical Research Institute of Jeonbuk National University Hospital, Jeon-ju 54907, Korea
- Correspondence: (D.H.L.); (H.S.K.); Tel.: +82-63-270-3998 (D.H.L.); +82-63-250-2582 (H.S.K.)
| |
Collapse
|
5
|
Chang C, Pan X, Tao H, Liu C, Veetil SP, Zhu J. 3D single-shot ptychography with highly tilted illuminations. OPTICS EXPRESS 2021; 29:30878-30891. [PMID: 34614805 DOI: 10.1364/oe.434613] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Received: 06/18/2021] [Accepted: 08/10/2021] [Indexed: 06/13/2023]
Abstract
A method based on highly tilted illumination and non-paraxial iterative computation is proposed to improve the image quality of single-shot 3D ptychography. A thick sample is illuminated with a cluster of laser beams that are separated by large enough angles to record each diffraction pattern distinctly in a single exposure. 3D structure of the thick sample is accurately reconstructed from recorded diffraction patterns using a modified multi-slice algorithm to process non-paraxial illumination. Sufficient number of recorded diffraction patterns with noticeably low crosstalk enhances the fidelity of reconstruction significantly over single-shot 3D ptychography methods that are based on paraxial illumination. Experimental observations guided by the results of numerical simulations show the feasibility of the proposed method.
Collapse
|
6
|
Urschel K, Tauchi M, Achenbach S, Dietel B. Investigation of Wall Shear Stress in Cardiovascular Research and in Clinical Practice-From Bench to Bedside. Int J Mol Sci 2021; 22:5635. [PMID: 34073212 PMCID: PMC8198948 DOI: 10.3390/ijms22115635] [Citation(s) in RCA: 23] [Impact Index Per Article: 7.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/02/2021] [Revised: 05/20/2021] [Accepted: 05/22/2021] [Indexed: 12/16/2022] Open
Abstract
In the 1900s, researchers established animal models experimentally to induce atherosclerosis by feeding them with a cholesterol-rich diet. It is now accepted that high circulating cholesterol is one of the main causes of atherosclerosis; however, plaque localization cannot be explained solely by hyperlipidemia. A tremendous amount of studies has demonstrated that hemodynamic forces modify endothelial athero-susceptibility phenotypes. Endothelial cells possess mechanosensors on the apical surface to detect a blood stream-induced force on the vessel wall, known as "wall shear stress (WSS)", and induce cellular and molecular responses. Investigations to elucidate the mechanisms of this process are on-going: on the one hand, hemodynamics in complex vessel systems have been described in detail, owing to the recent progress in imaging and computational techniques. On the other hand, investigations using unique in vitro chamber systems with various flow applications have enhanced the understanding of WSS-induced changes in endothelial cell function and the involvement of the glycocalyx, the apical surface layer of endothelial cells, in this process. In the clinical setting, attempts have been made to measure WSS and/or glycocalyx degradation non-invasively, for the purpose of their diagnostic utilization. An increasing body of evidence shows that WSS, as well as serum glycocalyx components, can serve as a predicting factor for atherosclerosis development and, most importantly, for the rupture of plaques in patients with high risk of coronary heart disease.
Collapse
Affiliation(s)
| | | | | | - Barbara Dietel
- Department of Medicine 2—Cardiology and Angiology, Friedrich-Alexander-University Erlangen-Nürnberg (FAU), Universitätsklinikum, 91054 Erlangen, Germany; (K.U.); (M.T.); (S.A.)
| |
Collapse
|
7
|
Wu W, Samant S, de Zwart G, Zhao S, Khan B, Ahmad M, Bologna M, Watanabe Y, Murasato Y, Burzotta F, Brilakis ES, Dangas G, Louvard Y, Stankovic G, Kassab GS, Migliavacca F, Chiastra C, Chatzizisis YS. 3D reconstruction of coronary artery bifurcations from coronary angiography and optical coherence tomography: feasibility, validation, and reproducibility. Sci Rep 2020; 10:18049. [PMID: 33093499 PMCID: PMC7582159 DOI: 10.1038/s41598-020-74264-w] [Citation(s) in RCA: 14] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/11/2020] [Accepted: 09/10/2020] [Indexed: 11/09/2022] Open
Abstract
The three-dimensional (3D) representation of the bifurcation anatomy and disease burden is essential for better understanding of the anatomical complexity of bifurcation disease and planning of stenting strategies. We propose a novel methodology for 3D reconstruction of coronary artery bifurcations based on the integration of angiography, which provides the backbone of the bifurcation, with optical coherence tomography (OCT), which provides the vessel shape. Our methodology introduces several technical novelties to tackle the OCT frame misalignment, correct positioning of the OCT frames at the carina, lumen surface reconstruction, and merging of bifurcation lumens. The accuracy and reproducibility of the methodology were tested in n = 5 patient-specific silicone bifurcations compared to contrast-enhanced micro-computed tomography (µCT), which was used as reference. The feasibility and time-efficiency of the method were explored in n = 7 diseased patient bifurcations of varying anatomical complexity. The OCT-based reconstructed bifurcation models were found to have remarkably high agreement compared to the µCT reference models, yielding r2 values between 0.91 and 0.98 for the normalized lumen areas, and mean differences of 0.005 for lumen shape and 0.004 degrees for bifurcation angles. Likewise, the reproducibility of our methodology was remarkably high. Our methodology successfully reconstructed all the patient bifurcations yielding favorable processing times (average lumen reconstruction time < 60 min). Overall, our method is an easily applicable, time-efficient, and user-friendly tool that allows accurate and reproducible 3D reconstruction of coronary bifurcations. Our technique can be used in the clinical setting to provide information about the bifurcation anatomy and plaque burden, thereby enabling planning, education, and decision making on bifurcation stenting.
Collapse
Affiliation(s)
- Wei Wu
- Cardiovasclar Biology and Biomechanics Laboratory, Cardiovascular Division, University of Nebraska Medical Center, Omaha, 68105, USA
| | - Saurabhi Samant
- Cardiovasclar Biology and Biomechanics Laboratory, Cardiovascular Division, University of Nebraska Medical Center, Omaha, 68105, USA
| | - Gijs de Zwart
- StudioGijs, Daendelsstraat 40, 5018 ES, Tilburg, The Netherlands
| | - Shijia Zhao
- Cardiovasclar Biology and Biomechanics Laboratory, Cardiovascular Division, University of Nebraska Medical Center, Omaha, 68105, USA
| | - Behram Khan
- Cardiovasclar Biology and Biomechanics Laboratory, Cardiovascular Division, University of Nebraska Medical Center, Omaha, 68105, USA
| | - Mansoor Ahmad
- Cardiovasclar Biology and Biomechanics Laboratory, Cardiovascular Division, University of Nebraska Medical Center, Omaha, 68105, USA
| | - Marco Bologna
- Biosignals, Bioimaging and Bioinformatics Laboratory (B3-Lab), Department of Electronics, Information and Bioengineering, Politecnico di Milano, 20133, Milan, Italy
| | - Yusuke Watanabe
- Department of Cardiology, Teikyo University Hospital, Tokyo, 173-0003, Japan
| | - Yoshinobu Murasato
- Department of Cardiology, National Hospital Organization Kyushu Medical Center, Fukuoka, 810-0065, Japan
| | - Francesco Burzotta
- Department of Cardiovascular Sciences, Fondazione Policlinico Universitario A. Gemelli IRCCS Università Cattolica del Sacro Cuore, 00168, Rome, Italy
| | | | - George Dangas
- Department of Cardiovascular Medicine, Mount Sinai Hospital, New York City, 10029, USA
| | - Yves Louvard
- Institut Cardiovasculaire Paris Sud, 91300, Massy, France
| | - Goran Stankovic
- Department of Cardiology, Clinical Center of Serbia, 11000, Belgrade, Serbia
| | - Ghassan S Kassab
- California Medical Innovation Institute, San Diego, CA, 92121, USA
| | - Francesco Migliavacca
- Laboratory of Biological Structure Mechanics (LaBS), Department of Chemistry, Materials and Chemical Engineering "Giulio Natta, Politecnico di Milano, 20133, Milan, Italy
| | - Claudio Chiastra
- PoliToBIOMed Lab, Department of Mechanical and Aerospace Engineering, Politecnico di Torino, 10129, Turin, Italy
| | - Yiannis S Chatzizisis
- Cardiovasclar Biology and Biomechanics Laboratory, Cardiovascular Division, University of Nebraska Medical Center, Omaha, 68105, USA.
| |
Collapse
|
8
|
Yang S, Kweon J, Roh JH, Lee JH, Kang H, Park LJ, Kim DJ, Yang H, Hur J, Kang DY, Lee PH, Ahn JM, Kang SJ, Park DW, Lee SW, Kim YH, Lee CW, Park SW, Park SJ. Deep learning segmentation of major vessels in X-ray coronary angiography. Sci Rep 2019; 9:16897. [PMID: 31729445 PMCID: PMC6858336 DOI: 10.1038/s41598-019-53254-7] [Citation(s) in RCA: 39] [Impact Index Per Article: 7.8] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/30/2019] [Accepted: 10/25/2019] [Indexed: 11/17/2022] Open
Abstract
X-ray coronary angiography is a primary imaging technique for diagnosing coronary diseases. Although quantitative coronary angiography (QCA) provides morphological information of coronary arteries with objective quantitative measures, considerable training is required to identify the target vessels and understand the tree structure of coronary arteries. Despite the use of computer-aided tools, such as the edge-detection method, manual correction is necessary for accurate segmentation of coronary vessels. In the present study, we proposed a robust method for major vessel segmentation using deep learning models with fully convolutional networks. When angiographic images of 3302 diseased major vessels from 2042 patients were tested, deep learning networks accurately identified and segmented the major vessels in X-ray coronary angiography. The average F1 score reached 0.917, and 93.7% of the images exhibited a high F1 score > 0.8. The most narrowed region at the stenosis was distinctly captured with high connectivity. Robust predictability was validated for the external dataset with different image characteristics. For major vessel segmentation, our approach demonstrated that prediction could be completed in real time with minimal image preprocessing. By applying deep learning segmentation, QCA analysis could be further automated, thereby facilitating the use of QCA-based diagnostic methods.
Collapse
Affiliation(s)
- Su Yang
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Jihoon Kweon
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea.
- Biomedical Engineering Research Center, Asan Medical Center, Seoul, Korea.
| | - Jae-Hyung Roh
- Department of Cardiology in Internal Medicine, School of Medicine, Chungnam National University, Chungnam National University Hospital, Daejeon, Korea
| | - Jae-Hwan Lee
- Department of Cardiology in Internal Medicine, School of Medicine, Chungnam National University, Chungnam National University Hospital, Daejeon, Korea
| | - Heejun Kang
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Lae-Jeong Park
- Department of Electronic Engineering, Gangneung-Wonju National University, Gangneung, Korea
| | - Dong Jun Kim
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Hyeonkyeong Yang
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Jaehee Hur
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Do-Yoon Kang
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Pil Hyung Lee
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Jung-Min Ahn
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Soo-Jin Kang
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Duk-Woo Park
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Seung-Whan Lee
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Young-Hak Kim
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea.
| | - Cheol Whan Lee
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Seong-Wook Park
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| | - Seung-Jung Park
- Division of Cardiology, Department of Internal Medicine, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
| |
Collapse
|