1
|
Xiong Z, Xia Q, Hu Z, Huang N, Bian C, Zheng Y, Vesal S, Ravikumar N, Maier A, Yang X, Heng PA, Ni D, Li C, Tong Q, Si W, Puybareau E, Khoudli Y, Géraud T, Chen C, Bai W, Rueckert D, Xu L, Zhuang X, Luo X, Jia S, Sermesant M, Liu Y, Wang K, Borra D, Masci A, Corsi C, de Vente C, Veta M, Karim R, Preetha CJ, Engelhardt S, Qiao M, Wang Y, Tao Q, Nuñez-Garcia M, Camara O, Savioli N, Lamata P, Zhao J. A global benchmark of algorithms for segmenting the left atrium from late gadolinium-enhanced cardiac magnetic resonance imaging. Med Image Anal 2021; 67:101832. [PMID: 33166776 DOI: 10.1016/j.media.2020.101832] [Citation(s) in RCA: 73] [Impact Index Per Article: 24.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/14/2020] [Revised: 09/21/2020] [Accepted: 09/23/2020] [Indexed: 12/29/2022]
Abstract
Segmentation of medical images, particularly late gadolinium-enhanced magnetic resonance imaging (LGE-MRI) used for visualizing diseased atrial structures, is a crucial first step for ablation treatment of atrial fibrillation. However, direct segmentation of LGE-MRIs is challenging due to the varying intensities caused by contrast agents. Since most clinical studies have relied on manual, labor-intensive approaches, automatic methods are of high interest, particularly optimized machine learning approaches. To address this, we organized the 2018 Left Atrium Segmentation Challenge using 154 3D LGE-MRIs, currently the world's largest atrial LGE-MRI dataset, and associated labels of the left atrium segmented by three medical experts, ultimately attracting the participation of 27 international teams. In this paper, extensive analysis of the submitted algorithms using technical and biological metrics was performed by undergoing subgroup analysis and conducting hyper-parameter analysis, offering an overall picture of the major design choices of convolutional neural networks (CNNs) and practical considerations for achieving state-of-the-art left atrium segmentation. Results show that the top method achieved a Dice score of 93.2% and a mean surface to surface distance of 0.7 mm, significantly outperforming prior state-of-the-art. Particularly, our analysis demonstrated that double sequentially used CNNs, in which a first CNN is used for automatic region-of-interest localization and a subsequent CNN is used for refined regional segmentation, achieved superior results than traditional methods and machine learning approaches containing single CNNs. This large-scale benchmarking study makes a significant step towards much-improved segmentation methods for atrial LGE-MRIs, and will serve as an important benchmark for evaluating and comparing the future works in the field. Furthermore, the findings from this study can potentially be extended to other imaging datasets and modalities, having an impact on the wider medical imaging community.
Collapse
Affiliation(s)
- Zhaohan Xiong
- Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand
| | - Qing Xia
- State Key Lab of Virtual Reality Technology and Systems, Beihang University, Beijing, China
| | - Zhiqiang Hu
- School of Electronics Engineering and Computer Science, Peking University, Beijing, China
| | | | - Cheng Bian
- Tencent Jarvis Laboratory, Shenzhen, China
| | | | - Sulaiman Vesal
- Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Nishant Ravikumar
- Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Andreas Maier
- Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Xin Yang
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong
| | - Pheng-Ann Heng
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong
| | - Dong Ni
- National-Regional Key Technology Engineering Laboratory for Medical Ultrasound, School of Biomedical Engineering, Health Science Center, Shenzhen University, Shenzhen, China
| | - Caizi Li
- School of Computer Science, Wuhan University, Wuhan, China
| | - Qianqian Tong
- School of Computer Science, Wuhan University, Wuhan, China
| | - Weixin Si
- Shenzhen Key Laboratory of Virtual Reality and Human Interaction Technology, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | | | - Younes Khoudli
- EPITA Research and Development Laboratory, Paris, France
| | - Thierry Géraud
- EPITA Research and Development Laboratory, Paris, France
| | - Chen Chen
- Department of Computing, Imperial College London, London, United Kingdom
| | - Wenjia Bai
- Department of Computing, Imperial College London, London, United Kingdom
| | - Daniel Rueckert
- Department of Computing, Imperial College London, London, United Kingdom
| | - Lingchao Xu
- School of Naval Architecture, Ocean & Civil Engineering, Shanghai Jiao Tong University, Shanghai, China
| | - Xiahai Zhuang
- School of Data Science, Fudan University, Shanghai, China
| | - Xinzhe Luo
- School of Data Science, Fudan University, Shanghai, China
| | - Shuman Jia
- Inria, Université Côte d'Azur, Epione team, Sophia Antipolis, France
| | - Maxime Sermesant
- Inria, Université Côte d'Azur, Epione team, Sophia Antipolis, France
| | - Yashu Liu
- School of Computer Science and Technology, Harbin Institute of Technology, Harbin, China
| | - Kuanquan Wang
- School of Computer Science and Technology, Harbin Institute of Technology, Harbin, China
| | - Davide Borra
- Department of Electric, Electronic and Information Engineering, University of Bologna, Cesena, Italy
| | - Alessandro Masci
- Department of Electric, Electronic and Information Engineering, University of Bologna, Cesena, Italy
| | - Cristiana Corsi
- Department of Electric, Electronic and Information Engineering, University of Bologna, Cesena, Italy
| | - Coen de Vente
- Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, the Netherlands
| | - Mitko Veta
- Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, the Netherlands
| | - Rashed Karim
- School of Biomedical Engineering & Imaging Sciences, Kings College London, London, United Kingdom
| | | | - Sandy Engelhardt
- Department of Internal Medicine III, Heidelberg University Hospital, Heidelberg, Germany
| | - Menyun Qiao
- Biomedical Engineering Center, Fudan University, Shanghai, China
| | - Yuanyuan Wang
- Biomedical Engineering Center, Fudan University, Shanghai, China
| | - Qian Tao
- Department of Radiology, Leiden University Medical Center, Leiden, the Netherlands
| | - Marta Nuñez-Garcia
- Physense, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain
| | - Oscar Camara
- Physense, Department of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain
| | - Nicolo Savioli
- Department of Bioengineering, Kings College London, London, United Kingdom
| | - Pablo Lamata
- Department of Bioengineering, Kings College London, London, United Kingdom
| | - Jichao Zhao
- Auckland Bioengineering Institute, University of Auckland, Auckland, New Zealand.
| |
Collapse
|
2
|
Duan J, Bello G, Schlemper J, Bai W, Dawes TJW, Biffi C, de Marvao A, Doumoud G, O'Regan DP, Rueckert D. Automatic 3D Bi-Ventricular Segmentation of Cardiac Images by a Shape-Refined Multi- Task Deep Learning Approach. IEEE TRANSACTIONS ON MEDICAL IMAGING 2019; 38:2151-2164. [PMID: 30676949 PMCID: PMC6728160 DOI: 10.1109/tmi.2019.2894322] [Citation(s) in RCA: 95] [Impact Index Per Article: 19.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/18/2023]
Abstract
Deep learning approaches have achieved state-of-the-art performance in cardiac magnetic resonance (CMR) image segmentation. However, most approaches have focused on learning image intensity features for segmentation, whereas the incorporation of anatomical shape priors has received less attention. In this paper, we combine a multi-task deep learning approach with atlas propagation to develop a shape-refined bi-ventricular segmentation pipeline for short-axis CMR volumetric images. The pipeline first employs a fully convolutional network (FCN) that learns segmentation and landmark localization tasks simultaneously. The architecture of the proposed FCN uses a 2.5D representation, thus combining the computational advantage of 2D FCNs networks and the capability of addressing 3D spatial consistency without compromising segmentation accuracy. Moreover, a refinement step is designed to explicitly impose shape prior knowledge and improve segmentation quality. This step is effective for overcoming image artifacts (e.g., due to different breath-hold positions and large slice thickness), which preclude the creation of anatomically meaningful 3D cardiac shapes. The pipeline is fully automated, due to network's ability to infer landmarks, which are then used downstream in the pipeline to initialize atlas propagation. We validate the pipeline on 1831 healthy subjects and 649 subjects with pulmonary hypertension. Extensive numerical experiments on the two datasets demonstrate that our proposed method is robust and capable of producing accurate, high-resolution, and anatomically smooth bi-ventricular 3D models, despite the presence of artifacts in input CMR volumes.
Collapse
|
3
|
Attar R, Pereañez M, Gooya A, Albà X, Zhang L, de Vila MH, Lee AM, Aung N, Lukaschuk E, Sanghvi MM, Fung K, Paiva JM, Piechnik SK, Neubauer S, Petersen SE, Frangi AF. Quantitative CMR population imaging on 20,000 subjects of the UK Biobank imaging study: LV/RV quantification pipeline and its evaluation. Med Image Anal 2019; 56:26-42. [PMID: 31154149 DOI: 10.1016/j.media.2019.05.006] [Citation(s) in RCA: 34] [Impact Index Per Article: 6.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/01/2018] [Revised: 05/13/2019] [Accepted: 05/23/2019] [Indexed: 01/02/2023]
Abstract
Population imaging studies generate data for developing and implementing personalised health strategies to prevent, or more effectively treat disease. Large prospective epidemiological studies acquire imaging for pre-symptomatic populations. These studies enable the early discovery of alterations due to impending disease, and enable early identification of individuals at risk. Such studies pose new challenges requiring automatic image analysis. To date, few large-scale population-level cardiac imaging studies have been conducted. One such study stands out for its sheer size, careful implementation, and availability of top quality expert annotation; the UK Biobank (UKB). The resulting massive imaging datasets (targeting ca. 100,000 subjects) has put published approaches for cardiac image quantification to the test. In this paper, we present and evaluate a cardiac magnetic resonance (CMR) image analysis pipeline that properly scales up and can provide a fully automatic analysis of the UKB CMR study. Without manual user interactions, our pipeline performs end-to-end image analytics from multi-view cine CMR images all the way to anatomical and functional bi-ventricular quantification. All this, while maintaining relevant quality controls of the CMR input images, and resulting image segmentations. To the best of our knowledge, this is the first published attempt to fully automate the extraction of global and regional reference ranges of all key functional cardiovascular indexes, from both left and right cardiac ventricles, for a population of 20,000 subjects imaged at 50 time frames per subject, for a total of one million CMR volumes. In addition, our pipeline provides 3D anatomical bi-ventricular models of the heart. These models enable the extraction of detailed information of the morphodynamics of the two ventricles for subsequent association to genetic, omics, lifestyle habits, exposure information, and other information provided in population imaging studies. We validated our proposed CMR analytics pipeline against manual expert readings on a reference cohort of 4620 subjects with contour delineations and corresponding clinical indexes. Our results show broad significant agreement between the manually obtained reference indexes, and those automatically computed via our framework. 80.67% of subjects were processed with mean contour distance of less than 1 pixel, and 17.50% with mean contour distance between 1 and 2 pixels. Finally, we compare our pipeline with a recently published approach reporting on UKB data, and based on deep learning. Our comparison shows similar performance in terms of segmentation accuracy with respect to human experts.
Collapse
Affiliation(s)
- Rahman Attar
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), School of Computing, University of Leeds, Leeds, UK; Biomedical Imaging Department, Leeds Institute for Cardiovascular and Metabolic Medicine (LICAMM), School of Medicine, University of Leeds, Leeds, UK.
| | - Marco Pereañez
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), School of Computing, University of Leeds, Leeds, UK; Biomedical Imaging Department, Leeds Institute for Cardiovascular and Metabolic Medicine (LICAMM), School of Medicine, University of Leeds, Leeds, UK
| | - Ali Gooya
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), School of Computing, University of Leeds, Leeds, UK
| | - Xènia Albà
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), Universitat Pompeu Fabra, Barcelona, Spain
| | - Le Zhang
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), Department of Electronic and Electrical Engineering, University of Sheffield, Sheffield, UK
| | - Milton Hoz de Vila
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), School of Computing, University of Leeds, Leeds, UK
| | - Aaron M Lee
- William Harvey Research Institute, NIHR Barts Biomedical Research Unit, Queen Mary University of London, London, UK; Barts Heart Centre, St Bartholomew's Hospital, Barts Health NHS Trust, London, UK
| | - Nay Aung
- William Harvey Research Institute, NIHR Barts Biomedical Research Unit, Queen Mary University of London, London, UK; Barts Heart Centre, St Bartholomew's Hospital, Barts Health NHS Trust, London, UK
| | - Elena Lukaschuk
- Oxford Centre for Clinical Magnetic Resonance Research (OCMR), Division of Cardiovascular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK
| | - Mihir M Sanghvi
- William Harvey Research Institute, NIHR Barts Biomedical Research Unit, Queen Mary University of London, London, UK; Barts Heart Centre, St Bartholomew's Hospital, Barts Health NHS Trust, London, UK
| | - Kenneth Fung
- William Harvey Research Institute, NIHR Barts Biomedical Research Unit, Queen Mary University of London, London, UK; Barts Heart Centre, St Bartholomew's Hospital, Barts Health NHS Trust, London, UK
| | - Jose Miguel Paiva
- William Harvey Research Institute, NIHR Barts Biomedical Research Unit, Queen Mary University of London, London, UK; Barts Heart Centre, St Bartholomew's Hospital, Barts Health NHS Trust, London, UK
| | - Stefan K Piechnik
- Oxford Centre for Clinical Magnetic Resonance Research (OCMR), Division of Cardiovascular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK
| | - Stefan Neubauer
- Oxford Centre for Clinical Magnetic Resonance Research (OCMR), Division of Cardiovascular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK
| | - Steffen E Petersen
- William Harvey Research Institute, NIHR Barts Biomedical Research Unit, Queen Mary University of London, London, UK; Barts Heart Centre, St Bartholomew's Hospital, Barts Health NHS Trust, London, UK
| | - Alejandro F Frangi
- Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), School of Computing, University of Leeds, Leeds, UK; Biomedical Imaging Department, Leeds Institute for Cardiovascular and Metabolic Medicine (LICAMM), School of Medicine, University of Leeds, Leeds, UK.
| |
Collapse
|
4
|
Multimodal imaging and three-dimensional cardiac computational modelling in the management of congenital heart disease: The secret to getting ahead is to get started. Arch Cardiovasc Dis 2018; 111:395-398. [DOI: 10.1016/j.acvd.2018.05.002] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/03/2018] [Revised: 05/15/2018] [Accepted: 05/15/2018] [Indexed: 11/23/2022]
|
5
|
Lim WK, Davila S, Teo JX, Yang C, Pua CJ, Blöcker C, Lim JQ, Ching J, Yap JJL, Tan SY, Sahlén A, Chin CWL, Teh BT, Rozen SG, Cook SA, Yeo KK, Tan P. Beyond fitness tracking: The use of consumer-grade wearable data from normal volunteers in cardiovascular and lipidomics research. PLoS Biol 2018; 16:e2004285. [PMID: 29485983 PMCID: PMC5828350 DOI: 10.1371/journal.pbio.2004285] [Citation(s) in RCA: 39] [Impact Index Per Article: 6.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/06/2017] [Accepted: 01/23/2018] [Indexed: 12/14/2022] Open
Abstract
The use of consumer-grade wearables for purposes beyond fitness tracking has not been comprehensively explored. We generated and analyzed multidimensional data from 233 normal volunteers, integrating wearable data, lifestyle questionnaires, cardiac imaging, sphingolipid profiling, and multiple clinical-grade cardiovascular and metabolic disease markers. We show that subjects can be stratified into distinct clusters based on daily activity patterns and that these clusters are marked by distinct demographic and behavioral patterns. While resting heart rates (RHRs) performed better than step counts in being associated with cardiovascular and metabolic disease markers, step counts identified relationships between physical activity and cardiac remodeling, suggesting that wearable data may play a role in reducing overdiagnosis of cardiac hypertrophy or dilatation in active individuals. Wearable-derived activity levels can be used to identify known and novel activity-modulated sphingolipids that are in turn associated with insulin sensitivity. Our findings demonstrate the potential for wearables in biomedical research and personalized health.
Collapse
Affiliation(s)
- Weng Khong Lim
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
- Cancer and Stem Biology Program, Duke-NUS Medical School, Singapore
| | - Sonia Davila
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
- Cardiovascular and Metabolic Disorders Program, Duke-NUS Medical School, Singapore
| | - Jing Xian Teo
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
| | - Chengxi Yang
- National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore
| | - Chee Jian Pua
- National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore
| | | | - Jing Quan Lim
- Lymphoma Genomic Translational Laboratory, Division of Medical Oncology, National Cancer Centre Singapore, Singapore
| | - Jianhong Ching
- Cardiovascular and Metabolic Disorders Program, Duke-NUS Medical School, Singapore
| | | | - Swee Yaw Tan
- Department of Cardiology, National Heart Centre Singapore, Singapore
| | - Anders Sahlén
- Department of Cardiology, National Heart Centre Singapore, Singapore
| | | | - Bin Tean Teh
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
- Cancer and Stem Biology Program, Duke-NUS Medical School, Singapore
- Laboratory of Cancer Epigenome, Division of Medical Sciences, National Cancer Centre Singapore, Singapore
- Institute of Molecular and Cell Biology, Agency for Science, Technology and Research, Singapore
- Cancer Science Institute of Singapore, National University of Singapore, Singapore
| | - Steven G. Rozen
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
- Cancer and Stem Biology Program, Duke-NUS Medical School, Singapore
- Centre for Computational Biology, Duke-NUS Medical School, Singapore
| | - Stuart Alexander Cook
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
- Cardiovascular and Metabolic Disorders Program, Duke-NUS Medical School, Singapore
- National Heart Research Institute Singapore, National Heart Centre Singapore, Singapore
- National Heart and Lung Institute, Imperial College London, United Kingdom
- MRC Clinical Sciences Centre, Imperial College London, United Kingdom
| | - Khung Keong Yeo
- Department of Cardiology, National Heart Centre Singapore, Singapore
| | - Patrick Tan
- SingHealth Duke-NUS Institute of Precision Medicine, Singapore
- Cancer and Stem Biology Program, Duke-NUS Medical School, Singapore
- Cancer Science Institute of Singapore, National University of Singapore, Singapore
- Biomedical Research Council, Agency for Science, Technology and Research, Singapore
| |
Collapse
|
6
|
Biffi C, de Marvao A, Attard MI, Dawes TJW, Whiffin N, Bai W, Shi W, Francis C, Meyer H, Buchan R, Cook SA, Rueckert D, O’Regan DP. Three-dimensional cardiovascular imaging-genetics: a mass univariate framework. Bioinformatics 2018; 34:97-103. [PMID: 28968671 PMCID: PMC5870605 DOI: 10.1093/bioinformatics/btx552] [Citation(s) in RCA: 21] [Impact Index Per Article: 3.5] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/05/2017] [Revised: 08/10/2017] [Accepted: 09/01/2017] [Indexed: 01/19/2023] Open
Abstract
Motivation Left ventricular (LV) hypertrophy is a strong predictor of cardiovascular outcomes, but its genetic regulation remains largely unexplained. Conventional phenotyping relies on manual calculation of LV mass and wall thickness, but advanced cardiac image analysis presents an opportunity for high-throughput mapping of genotype-phenotype associations in three dimensions (3D). Results High-resolution cardiac magnetic resonance images were automatically segmented in 1124 healthy volunteers to create a 3D shape model of the heart. Mass univariate regression was used to plot a 3D effect-size map for the association between wall thickness and a set of predictors at each vertex in the mesh. The vertices where a significant effect exists were determined by applying threshold-free cluster enhancement to boost areas of signal with spatial contiguity. Experiments on simulated phenotypic signals and SNP replication show that this approach offers a substantial gain in statistical power for cardiac genotype-phenotype associations while providing good control of the false discovery rate. This framework models the effects of genetic variation throughout the heart and can be automatically applied to large population cohorts. Availability and implementation The proposed approach has been coded in an R package freely available at https://doi.org/10.5281/zenodo.834610 together with the clinical data used in this work. Contact declan.oregan@imperial.ac.uk. Supplementary information Supplementary data are available at Bioinformatics online.
Collapse
Affiliation(s)
- Carlo Biffi
- Department of Computing, Imperial College London, South Kensington Campus, London, UK
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
| | - Antonio de Marvao
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
| | - Mark I Attard
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
| | - Timothy J W Dawes
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
- Quantitative Physiology and Genetics, National Heart and Lung Institute, Imperial College London, London, UK
| | - Nicola Whiffin
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
- Quantitative Physiology and Genetics, National Heart and Lung Institute, Imperial College London, London, UK
- NIHR Cardiovascular Biomedical Research Unit, Royal Brompton and Harefield NHS Trust, London, UK
| | - Wenjia Bai
- Department of Computing, Imperial College London, South Kensington Campus, London, UK
| | - Wenzhe Shi
- Department of Computing, Imperial College London, South Kensington Campus, London, UK
| | - Catherine Francis
- Quantitative Physiology and Genetics, National Heart and Lung Institute, Imperial College London, London, UK
- NIHR Cardiovascular Biomedical Research Unit, Royal Brompton and Harefield NHS Trust, London, UK
| | - Hannah Meyer
- European Molecular Biology Laboratory (EMBL), European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, UK
| | - Rachel Buchan
- Quantitative Physiology and Genetics, National Heart and Lung Institute, Imperial College London, London, UK
- NIHR Cardiovascular Biomedical Research Unit, Royal Brompton and Harefield NHS Trust, London, UK
| | - Stuart A Cook
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
- Quantitative Physiology and Genetics, National Heart and Lung Institute, Imperial College London, London, UK
- NIHR Cardiovascular Biomedical Research Unit, Royal Brompton and Harefield NHS Trust, London, UK
- Department of Cardiology, National Heart Centre Singapore, Singapore
- Programme in Cardiovascular and Metabolic Disorders, Duke National University Singapore, Singapore
| | - Daniel Rueckert
- Department of Computing, Imperial College London, South Kensington Campus, London, UK
| | - Declan P O’Regan
- Cardiovascular Magnetic Resonance Imaging and Genetics, MRC London Institute of Medical Sciences, Imperial College London, Hammersmith Hospital Campus, London, UK
| |
Collapse
|
7
|
Codari M, Scarabello M, Secchi F, Sforza C, Baselli G, Sardanelli F. Fully automated contour detection of the ascending aorta in cardiac 2D phase-contrast MRI. Magn Reson Imaging 2017; 47:77-82. [PMID: 29180100 DOI: 10.1016/j.mri.2017.11.010] [Citation(s) in RCA: 5] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/07/2017] [Revised: 11/15/2017] [Accepted: 11/21/2017] [Indexed: 11/16/2022]
Abstract
PURPOSE In this study we proposed a fully automated method for localizing and segmenting the ascending aortic lumen with phase-contrast magnetic resonance imaging (PC-MRI). MATERIAL AND METHODS Twenty-five phase-contrast series were randomly selected out of a large population dataset of patients whose cardiac MRI examination, performed from September 2008 to October 2013, was unremarkable. The local Ethical Committee approved this retrospective study. The ascending aorta was automatically identified on each phase of the cardiac cycle using a priori knowledge of aortic geometry. The frame that maximized the area, eccentricity, and solidity parameters was chosen for unsupervised initialization. Aortic segmentation was performed on each frame using active contouring without edges techniques. The entire algorithm was developed using Matlab R2016b. To validate the proposed method, the manual segmentation performed by a highly experienced operator was used. Dice similarity coefficient, Bland-Altman analysis, and Pearson's correlation coefficient were used as performance metrics. RESULTS Comparing automated and manual segmentation of the aortic lumen on 714 images, Bland-Altman analysis showed a bias of -6.68mm2, a coefficient of repeatability of 91.22mm2, a mean area measurement of 581.40mm2, and a reproducibility of 85%. Automated and manual segmentation were highly correlated (R=0.98). The Dice similarity coefficient versus the manual reference standard was 94.6±2.1% (mean±standard deviation). CONCLUSION A fully automated and robust method for identification and segmentation of ascending aorta on PC-MRI was developed. Its application on patients with a variety of pathologic conditions is advisable.
Collapse
Affiliation(s)
- Marina Codari
- Unit of Radiology, IRCCS Policlinico San Donato, Via Morandi 30, San Donato Milanese, 20097 Milan, Italy.
| | - Marco Scarabello
- Postgraduate School in Radiodiagnostics, Università degli Studi di Milano, Via Festa del Perdono 7, 20122 Milan, Italy
| | - Francesco Secchi
- Unit of Radiology, IRCCS Policlinico San Donato, Via Morandi 30, San Donato Milanese, 20097 Milan, Italy
| | - Chiarella Sforza
- Department of Biomedical Sciences for Health, Università degli Studi di Milano, Via Mangiagalli 31, Milan, 20133, Milan, Italy
| | - Giuseppe Baselli
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Piazza Leonardo Da Vinci 32, Milan, 20133, Milan, Italy
| | - Francesco Sardanelli
- Unit of Radiology, IRCCS Policlinico San Donato, Via Morandi 30, San Donato Milanese, 20097 Milan, Italy; Department of Biomedical Sciences for Health, Università degli Studi di Milano, Via Morandi 30, San Donato Milanese, 20097 Milan, Italy
| |
Collapse
|
8
|
Albà X, Lekadir K, Pereañez M, Medrano-Gracia P, Young AA, Frangi AF. Automatic initialization and quality control of large-scale cardiac MRI segmentations. Med Image Anal 2017; 43:129-141. [PMID: 29073531 DOI: 10.1016/j.media.2017.10.001] [Citation(s) in RCA: 31] [Impact Index Per Article: 4.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/17/2017] [Revised: 08/29/2017] [Accepted: 10/04/2017] [Indexed: 01/09/2023]
Abstract
Continuous advances in imaging technologies enable ever more comprehensive phenotyping of human anatomy and physiology. Concomitant reduction of imaging costs has resulted in widespread use of imaging in large clinical trials and population imaging studies. Magnetic Resonance Imaging (MRI), in particular, offers one-stop-shop multidimensional biomarkers of cardiovascular physiology and pathology. A wide range of analysis methods offer sophisticated cardiac image assessment and quantification for clinical and research studies. However, most methods have only been evaluated on relatively small databases often not accessible for open and fair benchmarking. Consequently, published performance indices are not directly comparable across studies and their translation and scalability to large clinical trials or population imaging cohorts is uncertain. Most existing techniques still rely on considerable manual intervention for the initialization and quality control of the segmentation process, becoming prohibitive when dealing with thousands of images. The contributions of this paper are three-fold. First, we propose a fully automatic method for initializing cardiac MRI segmentation, by using image features and random forests regression to predict an initial position of the heart and key anatomical landmarks in an MRI volume. In processing a full imaging database, the technique predicts the optimal corrective displacements and positions in relation to the initial rough intersections of the long and short axis images. Second, we introduce for the first time a quality control measure capable of identifying incorrect cardiac segmentations with no visual assessment. The method uses statistical, pattern and fractal descriptors in a random forest classifier to detect failures to be corrected or removed from subsequent statistical analysis. Finally, we validate these new techniques within a full pipeline for cardiac segmentation applicable to large-scale cardiac MRI databases. The results obtained based on over 1200 cases from the Cardiac Atlas Project show the promise of fully automatic initialization and quality control for population studies.
Collapse
Affiliation(s)
- Xènia Albà
- Center for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), Universitat Pompeu Fabra, Barcelona, Spain.
| | - Karim Lekadir
- Center for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), Universitat Pompeu Fabra, Barcelona, Spain
| | - Marco Pereañez
- Center for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), Electronic and Electrical Engineering Department, University of Sheffield, Sheffield, UK
| | - Pau Medrano-Gracia
- Department of Anatomy and Medical Imaging, University of Auckland, Auckland, NZ
| | - Alistair A Young
- Department of Anatomy and Medical Imaging, University of Auckland, Auckland, NZ
| | - Alejandro F Frangi
- Center for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), Electronic and Electrical Engineering Department, University of Sheffield, Sheffield, UK
| |
Collapse
|
9
|
Suinesiaputra A, Sanghvi MM, Aung N, Paiva JM, Zemrak F, Fung K, Lukaschuk E, Lee AM, Carapella V, Kim YJ, Francis J, Piechnik SK, Neubauer S, Greiser A, Jolly MP, Hayes C, Young AA, Petersen SE. Fully-automated left ventricular mass and volume MRI analysis in the UK Biobank population cohort: evaluation of initial results. Int J Cardiovasc Imaging 2017; 34:281-291. [PMID: 28836039 PMCID: PMC5809564 DOI: 10.1007/s10554-017-1225-9] [Citation(s) in RCA: 34] [Impact Index Per Article: 4.9] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 04/28/2017] [Accepted: 08/05/2017] [Indexed: 12/26/2022]
Abstract
UK Biobank, a large cohort study, plans to acquire 100,000 cardiac MRI studies by 2020. Although fully-automated left ventricular (LV) analysis was performed in the original acquisition, this was not designed for unsupervised incorporation into epidemiological studies. We sought to evaluate automated LV mass and volume (Siemens syngo InlineVF versions D13A and E11C), against manual analysis in a substantial sub-cohort of UK Biobank participants. Eight readers from two centers, trained to give consistent results, manually analyzed 4874 UK Biobank cases for LV end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV), ejection fraction (EF) and LV mass (LVM). Agreement between manual and InlineVF automated analyses were evaluated using Bland–Altman analysis and the intra-class correlation coefficient (ICC). Tenfold cross-validation was used to establish a linear regression calibration between manual and InlineVF results. InlineVF D13A returned results in 4423 cases, whereas InlineVF E11C returned results in 4775 cases and also reported LVM. Rapid visual assessment of the E11C results found 178 cases (3.7%) with grossly misplaced contours or landmarks. In the remaining 4597 cases, LV function showed good agreement: ESV −6.4 ± 9.0 ml, 0.853 (mean ± SD of the differences, ICC) EDV −3.0 ± 11.6 ml, 0.937; SV 3.4 ± 9.8 ml, 0.855; and EF 3.5 ± 5.1%, 0.586. Although LV mass was consistently overestimated (29.9 ± 17.0 g, 0.534) due to larger epicardial contours on all slices, linear regression could be used to correct the bias and improve accuracy. Automated InlineVF results can be used for case-control studies in UK Biobank, provided visual quality control and linear bias correction are performed. Improvements between InlineVF D13A and InlineVF E11C show the field is rapidly advancing, with further improvements expected in the near future.
Collapse
Affiliation(s)
- Avan Suinesiaputra
- Department of Anatomy and Medical Imaging, Faculty of Medical and Health Sciences, University of Auckland, 85 Park Road, Auckland, 1142 New Zealand
| | - Mihir M. Sanghvi
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| | - Nay Aung
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| | - Jose Miguel Paiva
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| | - Filip Zemrak
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| | - Kenneth Fung
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| | - Elena Lukaschuk
- Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, UK
| | - Aaron M. Lee
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| | - Valentina Carapella
- Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, UK
| | - Young Jin Kim
- Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, UK
| | - Jane Francis
- Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, UK
| | - Stefan K. Piechnik
- Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, UK
| | - Stefan Neubauer
- Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford, Oxford, UK
| | | | | | | | - Alistair A. Young
- Department of Anatomy and Medical Imaging, Faculty of Medical and Health Sciences, University of Auckland, 85 Park Road, Auckland, 1142 New Zealand
| | - Steffen E. Petersen
- William Harvey Research Institute, NIHR Cardiovascular Biomedical Research Centre at Barts, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ UK
| |
Collapse
|
10
|
Bruse JL, Zuluaga MA, Khushnood A, McLeod K, Ntsinjana HN, Hsia TY, Sermesant M, Pennec X, Taylor AM, Schievano S. Detecting Clinically Meaningful Shape Clusters in Medical Image Data: Metrics Analysis for Hierarchical Clustering Applied to Healthy and Pathological Aortic Arches. IEEE Trans Biomed Eng 2017; 64:2373-2383. [PMID: 28221991 DOI: 10.1109/tbme.2017.2655364] [Citation(s) in RCA: 31] [Impact Index Per Article: 4.4] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/09/2022]
Abstract
OBJECTIVE Today's growing medical image databases call for novel processing tools to structure the bulk of data and extract clinically relevant information. Unsupervised hierarchical clustering may reveal clusters within anatomical shape data of patient populations as required for modern precision medicine strategies. Few studies have applied hierarchical clustering techniques to three-dimensional patient shape data and results depend heavily on the chosen clustering distance metrics and linkage functions. In this study, we sought to assess clustering classification performance of various distance/linkage combinations and of different types of input data to obtain clinically meaningful shape clusters. METHODS We present a processing pipeline combining automatic segmentation, statistical shape modeling, and agglomerative hierarchical clustering to automatically subdivide a set of 60 aortic arch anatomical models into healthy controls, two groups affected by congenital heart disease, and their respective subgroups as defined by clinical diagnosis. Results were compared with traditional morphometrics and principal component analysis of shape features. RESULTS Our pipeline achieved automatic division of input shape data according to primary clinical diagnosis with high F-score (0.902 ± 0.042) and Matthews correlation coefficient (0.851 ± 0.064) using the correlation/weighted distance/linkage combination. Meaningful subgroups within the three patient groups were obtained and benchmark scores for automatic segmentation and classification performance are reported. CONCLUSION Clustering results vary depending on the distance/linkage combination used to divide the data. Yet, clinically relevant shape clusters and subgroups could be found with high specificity and low misclassification rates. SIGNIFICANCE Detecting disease-specific clusters within medical image data could improve image-based risk assessment, treatment planning, and medical device development in complex disease.
Collapse
|
11
|
Frangi AF, Taylor ZA, Gooya A. Precision Imaging: more descriptive, predictive and integrative imaging. Med Image Anal 2016; 33:27-32. [PMID: 27373145 DOI: 10.1016/j.media.2016.06.024] [Citation(s) in RCA: 7] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/30/2016] [Revised: 06/15/2016] [Accepted: 06/15/2016] [Indexed: 12/22/2022]
Abstract
Medical image analysis has grown into a matured field challenged by progress made across all medical imaging technologies and more recent breakthroughs in biological imaging. The cross-fertilisation between medical image analysis, biomedical imaging physics and technology, and domain knowledge from medicine and biology has spurred a truly interdisciplinary effort that stretched outside the original boundaries of the disciplines that gave birth to this field and created stimulating and enriching synergies. Consideration on how the field has evolved and the experience of the work carried out over the last 15 years in our centre, has led us to envision a future emphasis of medical imaging in Precision Imaging. Precision Imaging is not a new discipline but rather a distinct emphasis in medical imaging borne at the cross-roads between, and unifying the efforts behind mechanistic and phenomenological model-based imaging. It captures three main directions in the effort to deal with the information deluge in imaging sciences, and thus achieve wisdom from data, information, and knowledge. Precision Imaging is finally characterised by being descriptive, predictive and integrative about the imaged object. This paper provides a brief and personal perspective on how the field has evolved, summarises and formalises our vision of Precision Imaging for Precision Medicine, and highlights some connections with past research and current trends in the field.
Collapse
Affiliation(s)
- Alejandro F Frangi
- CISTIB Centre for Computational Imaging & Simulation Technologies in Biomedicine, Electronic and Electrical Engineering Department, University of Sheffield, Sheffield, UK.
| | - Zeike A Taylor
- CISTIB Centre for Computational Imaging & Simulation Technologies in Biomedicine, Mechanical Engineering Department, University of Sheffield, Sheffield, UK.
| | - Ali Gooya
- CISTIB Centre for Computational Imaging & Simulation Technologies in Biomedicine, Electronic and Electrical Engineering Department, University of Sheffield, Sheffield, UK.
| |
Collapse
|