6
|
Zhang N, Yang G, Gao Z, Xu C, Zhang Y, Shi R, Keegan J, Xu L, Zhang H, Fan Z, Firmin D. Deep Learning for Diagnosis of Chronic Myocardial Infarction on Nonenhanced Cardiac Cine MRI. Radiology 2019; 291:606-617. [PMID: 31038407 DOI: 10.1148/radiol.2019182304] [Citation(s) in RCA: 109] [Impact Index Per Article: 21.8] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/17/2023]
Abstract
Background Renal impairment is common in patients with coronary artery disease and, if severe, late gadolinium enhancement (LGE) imaging for myocardial infarction (MI) evaluation cannot be performed. Purpose To develop a fully automatic framework for chronic MI delineation via deep learning on non-contrast material-enhanced cardiac cine MRI. Materials and Methods In this retrospective single-center study, a deep learning model was developed to extract motion features from the left ventricle and delineate MI regions on nonenhanced cardiac cine MRI collected between October 2015 and March 2017. Patients with chronic MI, as well as healthy control patients, had both nonenhanced cardiac cine (25 phases per cardiac cycle) and LGE MRI examinations. Eighty percent of MRI examinations were used for the training data set and 20% for the independent testing data set. Chronic MI regions on LGE MRI were defined as ground truth. Diagnostic performance was assessed by analysis of the area under the receiver operating characteristic curve (AUC). MI area and MI area percentage from nonenhanced cardiac cine and LGE MRI were compared by using the Pearson correlation, paired t test, and Bland-Altman analysis. Results Study participants included 212 patients with chronic MI (men, 171; age, 57.2 years ± 12.5) and 87 healthy control patients (men, 42; age, 43.3 years ± 15.5). Using the full cardiac cine MRI, the per-segment sensitivity and specificity for detecting chronic MI in the independent test set was 89.8% and 99.1%, respectively, with an AUC of 0.94. There were no differences between nonenhanced cardiac cine and LGE MRI analyses in number of MI segments (114 vs 127, respectively; P = .38), per-patient MI area (6.2 cm2 ± 2.8 vs 5.5 cm2 ± 2.3, respectively; P = .27; correlation coefficient, r = 0.88), and MI area percentage (21.5% ± 17.3 vs 18.5% ± 15.4; P = .17; correlation coefficient, r = 0.89). Conclusion The proposed deep learning framework on nonenhanced cardiac cine MRI enables the confirmation (presence), detection (position), and delineation (transmurality and size) of chronic myocardial infarction. However, future larger-scale multicenter studies are required for a full validation. Published under a CC BY 4.0 license. Online supplemental material is available for this article. See also the editorial by Leiner in this issue.
Collapse
Affiliation(s)
- Nan Zhang
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Guang Yang
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Zhifan Gao
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Chenchu Xu
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Yanping Zhang
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Rui Shi
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Jennifer Keegan
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Lei Xu
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Heye Zhang
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - Zhanming Fan
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| | - David Firmin
- From the Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, 2nd Anzhen Road, Chaoyang District, Beijing, China (N.Z., L.X., Z.F.); Cardiovascular Research Centre, Royal Brompton Hospital, London, England (G.Y., R.S., J.K., D.F.); National Heart and Lung Institute, Imperial College London, London, England (G.Y., R.S., J.K., D.F.); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China (Z.G., H.Z.); Anhui University, Hefei, China (C.X., Y.Z.); and School of Biomedical Engineering, Sun Yat-Sen University, Shenzhen, China (H.Z.)
| |
Collapse
|
8
|
Zhang L, Huttin O, Marie PY, Felblinger J, Beaumont M, Chillou CDE, Girerd N, Mandry D. Myocardial infarct sizing by late gadolinium-enhanced MRI: Comparison of manual, full-width at half-maximum, and n-standard deviation methods. J Magn Reson Imaging 2016; 44:1206-1217. [PMID: 27096741 DOI: 10.1002/jmri.25285] [Citation(s) in RCA: 14] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/21/2015] [Accepted: 03/31/2016] [Indexed: 11/11/2022] Open
Abstract
PURPOSE To compare three widely used methods for myocardial infarct (MI) sizing on late gadolinium-enhanced (LGE) magnetic resonance (MR) images: manual delineation and two semiautomated techniques (full-width at half-maximum [FWHM] and n-standard deviation [SD]). MATERIALS AND METHODS 3T phase-sensitive inversion-recovery (PSIR) LGE images of 114 patients after an acute MI (2-4 days and 6 months) were analyzed by two independent observers to determine both total and core infarct sizes (TIS/CIS). Manual delineation served as the reference for determination of optimal thresholds for semiautomated methods after thresholding at multiple values. Reproducibility and accuracy were expressed as overall bias ± 95% limits of agreement. RESULTS Mean infarct sizes by manual methods were 39.0%/24.4% for the acute MI group (TIS/CIS) and 29.7%/17.3% for the chronic MI group. The optimal thresholds (ie, providing the closest mean value to the manual method) were FWHM30% and 3SD for the TIS measurement and FWHM45% and 6SD for the CIS measurement (paired t-test; all P > 0.05). The best reproducibility was obtained using FWHM. For TIS measurement in the acute MI group, intra-/interobserver agreements, from Bland-Altman analysis, with FWHM30%, 3SD, and manual were -0.02 ± 7.74%/-0.74 ± 5.52%, 0.31 ± 9.78%/2.96 ± 16.62% and -2.12 ± 8.86%/0.18 ± 16.12, respectively; in the chronic MI group, the corresponding values were 0.23 ± 3.5%/-2.28 ± 15.06, -0.29 ± 10.46%/3.12 ± 13.06% and 1.68 ± 6.52%/-2.88 ± 9.62%, respectively. A similar trend for reproducibility was obtained for CIS measurement. However, semiautomated methods produced inconsistent results (variabilities of 24-46%) compared to manual delineation. CONCLUSION The FWHM technique was the most reproducible method for infarct sizing both in acute and chronic MI. However, both FWHM and n-SD methods showed limited accuracy compared to manual delineation. J. Magn. Reson. Imaging 2016;44:1206-1217.
Collapse
Affiliation(s)
- Lin Zhang
- INSERM, U947, IADI, Nancy, F-54000, France.,Université de Lorraine, Nancy, F-54000, France
| | - Olivier Huttin
- CHRU Nancy, Departement de Cardiologie, Nancy, F-54000, France
| | - Pierre-Yves Marie
- Université de Lorraine, Nancy, F-54000, France.,INSERM, U961, Nancy, F-54000, France.,CHRU Nancy, Pôle Imagerie, Nancy, F-54000, France
| | - Jacques Felblinger
- INSERM, U947, IADI, Nancy, F-54000, France.,Université de Lorraine, Nancy, F-54000, France.,CHRU Nancy, Pôle Imagerie, Nancy, F-54000, France.,INSERM, CIC-IT 1433, Nancy, F-54000, France
| | - Marine Beaumont
- INSERM, U947, IADI, Nancy, F-54000, France.,INSERM, CIC-IT 1433, Nancy, F-54000, France
| | - Christian DE Chillou
- INSERM, U947, IADI, Nancy, F-54000, France.,Université de Lorraine, Nancy, F-54000, France.,CHRU Nancy, Departement de Cardiologie, Nancy, F-54000, France
| | - Nicolas Girerd
- Université de Lorraine, Nancy, F-54000, France.,CHRU Nancy, Departement de Cardiologie, Nancy, F-54000, France.,INSERM, CIC-P 9501, Nancy, F-54000, France
| | - Damien Mandry
- INSERM, U947, IADI, Nancy, F-54000, France. .,Université de Lorraine, Nancy, F-54000, France. .,CHRU Nancy, Pôle Imagerie, Nancy, F-54000, France.
| |
Collapse
|
13
|
Huang Z, Li C, Yang S, Xu J, Shen Y, Xie X, Dai Y, Lu H, Gong H, Sun A, Qian J, Ge J. Magnetic resonance hypointensive signal primarily originates from extracellular iron particles in the long-term tracking of mesenchymal stem cells transplanted in the infarcted myocardium. Int J Nanomedicine 2015. [PMID: 25767388 PMCID: PMC4354691 DOI: 10.2147/ijn.s77858] [Citation(s) in RCA: 18] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/23/2022] Open
Abstract
PURPOSE The long-lasting hypointensities in cardiac magnetic resonance (CMR) were believed to originate from superparamagnetic iron oxide (SPIO)-engulfed macrophages during long-term stem cell tracking. However, the iron clearance capacity of the ischemic heart was limited. Therefore, we speculated that the extracellular SPIO particles may also be involved in the generation of false-positive signals. METHODS AND RESULTS Male swine mesenchymal stem cells (MSCs) were incubated with SPIO for 24 hours, and SPIO labeling had no significant effects on either cell viability or differentiation. In vitro studies showed that magnetic resonance failed to distinguish SPIO from living SPIO-MSCs or dead SPIO-MSCs. Two hours after the establishment of the female swine acute myocardial infarction model, 2×10(7) male SPIO-labeled MSCs (n=5) or unlabeled MSCs (n=5) were transextracardially injected into the infarcted myocardium at ten distinct sites. In vivo CMR with T2 star weighted imaging-flash-2D sequence revealed a signal void corresponding to the initial SPIO-MSC injection sites. At 6 months after transplantation, CMR identified 32 (64%) of the 50 injection sites, where massive Prussian blue-positive iron deposits were detected by pathological examination. However, iron particles were predominantly distributed in the extracellular space, and a minority was distributed within CD68-positive macrophages and other CD68-negative cells. No sex-determining region Y DNA of donor MSCs was detected. CONCLUSION CMR hypointensive signal is primarily caused by extracellular iron particles in the long-term tracking of transplanted MSCs after myocardial infarction. Consideration should be given to both the false-positive signal and the potential cardiac toxicity of long-standing iron deposits in the heart.
Collapse
Affiliation(s)
- Zheyong Huang
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Chenguang Li
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Shan Yang
- Department of Radiology, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Jianfeng Xu
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Yunli Shen
- Department of Cardiology, Shanghai East Hospital, Tongji University, Shanghai, People's Republic of China
| | - Xinxing Xie
- Department of Cardiology, Qianfoshan Hospital, Shandong University, Jinan, Shandong Province, People's Republic of China
| | - Yuxiang Dai
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Hao Lu
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Hui Gong
- Institute of Biomedical Science, Fudan University, Shanghai, People's Republic of China
| | - Aijun Sun
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Juying Qian
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| | - Junbo Ge
- Shanghai Institute of Cardiovascular Diseases, Zhongshan Hospital, Fudan University, Shanghai, People's Republic of China
| |
Collapse
|