Deep Learning Synthetic Strain: Quantitative Assessment of Regional Myocardial Wall Motion at MRI

Published Online:May 11 2023https://doi.org/10.1148/ryct.220202

Abstract

A deep learning algorithm was able to infer myocardial velocities and quantify strain from cine steady-state free precession images to detect wall motion abnormalities in patients with ischemic heart disease, performing comparably with subspecialty radiologists.

Purpose

To assess the feasibility of a newly developed algorithm, called deep learning synthetic strain (DLSS), to infer myocardial velocity from cine steady-state free precession (SSFP) images and detect wall motion abnormalities in patients with ischemic heart disease.

Materials and Methods

In this retrospective study, DLSS was developed by using a data set of 223 cardiac MRI examinations including cine SSFP images and four-dimensional flow velocity data (November 2017 to May 2021). To establish normal ranges, segmental strain was measured in 40 individuals (mean age, 41 years ± 17 [SD]; 30 men) without cardiac disease. Then, DLSS performance in the detection of wall motion abnormalities was assessed in a separate group of patients with coronary artery disease, and these findings were compared with consensus results of four independent cardiothoracic radiologists (ground truth). Algorithm performance was evaluated by using receiver operating characteristic curve analysis.

Results

Median peak segmental radial strain in individuals with normal cardiac MRI findings was 38% (IQR: 30%–48%). Among patients with ischemic heart disease (846 segments in 53 patients; mean age, 61 years ± 12; 41 men), the Cohen κ among four cardiothoracic readers for detecting wall motion abnormalities was 0.60–0.78. DLSS achieved an area under the receiver operating characteristic curve of 0.90. Using a fixed 30% threshold for abnormal peak radial strain, the algorithm achieved a sensitivity, specificity, and accuracy of 86%, 85%, and 86%, respectively.

Conclusion

The deep learning algorithm had comparable performance with subspecialty radiologists in inferring myocardial velocity from cine SSFP images and identifying myocardial wall motion abnormalities at rest in patients with ischemic heart disease.

Keywords: Neural Networks, Cardiac, MR Imaging, Ischemia/Infarction

Supplemental material is available for this article.

References

1. Rose MJ, Jarvis K, Chowdhary V, et al. Efficient method for volumetric assessment of peak blood flow velocity using 4D flow MRI. J Magn Reson Imaging 2016;44(6):1673–1682. Crossref, Medline, Google Scholar
2. Malayeri AA, Johnson WC, Macedo R, Bathon J, Lima JAC, Bluemke DA. Cardiac cine MRI: quantification of the relationship between fast gradient echo and steady-state free precession for determination of myocardial mass and volumes. J Magn Reson Imaging 2008;28(1):60–66. Crossref, Medline, Google Scholar
3. La Gerche A, Claessen G, Van de Bruaene A, et al. Cardiac MRI: a new gold standard for ventricular volume quantification during high-intensity exercise. Circ Cardiovasc Imaging 2013;6(2):329–338. Crossref, Medline, Google Scholar
4. Koch R, Lang RM, Garcia MJ, et al. Objective evaluation of regional left ventricular wall motion during dobutamine stress echocardiographic studies using segmental analysis of color kinesis images. J Am Coll Cardiol 1999;34(2):409–419. Crossref, Medline, Google Scholar
5. Sharir T, Berman DS, Waechter PB, et al. Quantitative analysis of regional motion and thickening by gated myocardial perfusion SPECT: normal heterogeneity and criteria for abnormality. J Nucl Med 2001;42(11):1630–1638. Medline, Google Scholar
6. Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18(12):1440–1463. Crossref, Medline, Google Scholar
7. Flachskampf FA, Schmid M, Rost C, Achenbach S, DeMaria AN, Daniel WG. Cardiac imaging after myocardial infarction. Eur Heart J 2011;32(3):272–283. Crossref, Medline, Google Scholar
8. West AM, Kramer CM. Cardiovascular magnetic resonance imaging of myocardial infarction, viability, and cardiomyopathies. Curr Probl Cardiol 2010;35(4):176–220. Crossref, Medline, Google Scholar
9. Jahnke C, Nagel E, Gebker R, et al. Prognostic value of cardiac magnetic resonance stress tests: adenosine stress perfusion and dobutamine stress wall motion imaging. Circulation 2007;115(13):1769–1776. Crossref, Medline, Google Scholar
10. Roditi GH, Hartnell GG, Cohen MC. MRI changes in myocarditis—evaluation with spin echo, cine MR angiography and contrast enhanced spin echo imaging. Clin Radiol 2000;55(10):752–758. Crossref, Medline, Google Scholar
11. Gandjbakhch E, Redheuil A, Pousset F, Charron P, Frank R. Clinical diagnosis, imaging, and genetics of arrhythmogenic right ventricular cardiomyopathy/dysplasia: JACC state-of-the-art review. J Am Coll Cardiol 2018;72(7):784–804. Crossref, Medline, Google Scholar
12. Zerhouni EA, Parish DM, Rogers WJ, Yang A, Shapiro EP. Human heart: tagging with MR imaging—a method for noninvasive assessment of myocardial motion. Radiology 1988;169(1):59–63. Link, Google Scholar
13. Auger DA, Zhong X, Epstein FH, Spottiswoode BS. Mapping right ventricular myocardial mechanics using 3D cine DENSE cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2012;14(1):4. Crossref, Medline, Google Scholar
14. Korosoglou G, Gitsioudis G, Voss A, et al. Strain-encoded cardiac magnetic resonance during high-dose dobutamine stress testing for the estimation of cardiac outcomes: comparison to clinical parameters and conventional wall motion readings. J Am Coll Cardiol 2011;58(11):1140–1149. Crossref, Medline, Google Scholar
15. Osman NF, Kerwin WS, McVeigh ER, Prince JL. Cardiac motion tracking using CINE harmonic phase (HARP) magnetic resonance imaging. Magn Reson Med 1999;42(6):1048–1060. Crossref, Medline, Google Scholar
16. Ruh A, Sarnari R, Berhane H, et al. Impact of age and cardiac disease on regional left and right ventricular myocardial motion in healthy controls and patients with repaired tetralogy of Fallot. Int J Cardiovasc Imaging 2019;35(6):1119–1132. Crossref, Medline, Google Scholar
17. Arai AE, Gaither CC 3rd, Epstein FH, Balaban RS, Wolff SD. Myocardial velocity gradient imaging by phase contrast MRI with application to regional function in myocardial ischemia. Magn Reson Med 1999;42(1):98–109. Crossref, Medline, Google Scholar
18. Maret E, Todt T, Brudin L, et al. Functional measurements based on feature tracking of cine magnetic resonance images identify left ventricular segments with myocardial scar. Cardiovasc Ultrasound 2009;7(1):53. Crossref, Medline, Google Scholar
19. Backhaus SJ, Metschies G, Zieschang V, et al. Head-to-head comparison of cardiovascular MR feature tracking cine versus acquisition-based deformation strain imaging using myocardial tagging and strain encoding. Magn Reson Med 2021;85(1):357–368. Crossref, Medline, Google Scholar
20. Scatteia A, Baritussio A, Bucciarelli-Ducci C. Strain imaging using cardiac magnetic resonance. Heart Fail Rev 2017;22(4):465–476. Crossref, Medline, Google Scholar
21. Mirea O, Pagourelias ED, Duchenne J, et al. Intervendor differences in the accuracy of detecting regional functional abnormalities: a report from the EACVI-ASE strain standardization task force. JACC Cardiovasc Imaging 2018;11(1):25–34. Crossref, Medline, Google Scholar
22. Popović ZB, Benejam C, Bian J, et al. Speckle-tracking echocardiography correctly identifies segmental left ventricular dysfunction induced by scarring in a rat model of myocardial infarction. Am J Physiol Heart Circ Physiol 2007;292(6):H2809–H2816. Crossref, Medline, Google Scholar
23. Roes SD, Mollema SA, Lamb HJ, van der Wall EE, de Roos A, Bax JJ. Validation of echocardiographic two-dimensional speckle tracking longitudinal strain imaging for viability assessment in patients with chronic ischemic left ventricular dysfunction and comparison with contrast-enhanced magnetic resonance imaging. Am J Cardiol 2009;104(3):312–317. Crossref, Medline, Google Scholar
24. Vo HQ, Marwick TH, Negishi K. MRI-derived myocardial strain measures in normal subjects. JACC Cardiovasc Imaging 2018;11(2 Pt 1):196–205. Crossref, Medline, Google Scholar
25. Bucius P, Erley J, Tanacli R, et al. Comparison of feature tracking, fast-SENC, and myocardial tagging for global and segmental left ventricular strain. ESC Heart Fail 2020;7(2):523–532. Crossref, Medline, Google Scholar
26. Wu L, Germans T, Güçlü A, Heymans MW, Allaart CP, van Rossum AC. Feature tracking compared with tissue tagging measurements of segmental strain by cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2014;16(1):10. Crossref, Medline, Google Scholar
27. Fries JA, Varma P, Chen VS, et al. Weakly supervised classification of aortic valve malformations using unlabeled cardiac MRI sequences. Nat Commun 2019;10(1):3111. Crossref, Medline, Google Scholar
28. Xue H, Artico J, Fontana M, Moon JC, Davies RH, Kellman P. Landmark detection in cardiac MRI by using a convolutional neural network. Radiol Artif Intell 2021;3(5):e200197. Link, Google Scholar
29. Blansit K, Retson T, Masutani E, Bahrami N, Hsiao A. Deep learning-based prescription of cardiac MRI planes. Radiol Artif Intell 2019;1(6):e180069. Link, Google Scholar
30. Zotti C, Luo Z, Lalande A, Jodoin PM. Convolutional neural network with shape prior applied to cardiac MRI segmentation. IEEE J Biomed Health Inform 2019;23(3):1119–1128. Crossref, Medline, Google Scholar
31. Lieman-Sifry J, Le M, Lau F, Sall S, Golden D. FastVentricle: cardiac segmentation with ENet. In: Pop M, Wright G, eds. Functional imaging and modelling of the heart. FIMH 2017. Lecture Notes in Computer Science, vol 10263. Springer, 2017; 127–138. Crossref, Google Scholar
32. Tao Q, Yan W, Wang Y, et al. Deep learning-based method for fully automatic quantification of left ventricle function from cine MR images: a multivendor, multicenter study. Radiology 2019;290(1):81–88. Link, Google Scholar
33. Masutani EM, Bahrami N, Hsiao A. Deep learning single-frame and multiframe super-resolution for cardiac MRI. Radiology 2020;295(3):552–561. Link, Google Scholar
34. You S, Masutani EM, Alley MT, et al. Deep learning automated background phase error correction for abdominopelvic 4D flow MRI. Radiology 2022;302(3):584–592. Link, Google Scholar
35. Ghadimi S, Auger DA, Feng X, et al. Fully-automated global and segmental strain analysis of DENSE cardiovascular magnetic resonance using deep learning for segmentation and phase unwrapping. J Cardiovasc Magn Reson 2021;23(1):20. Crossref, Medline, Google Scholar
36. Ferdian E, Suinesiaputra A, Fung K, et al. Fully automated myocardial strain estimation from cardiovascular MRI-tagged images using a deep learning framework in the UK biobank. Radiol Cardiothorac Imaging 2020;2(1):e190032. Link, Google Scholar
37. Retson TA, Masutani EM, Golden D, Hsiao A. Clinical performance and role of expert supervision of deep learning for cardiac ventricular volumetry: a validation study. Radiol Artif Intell 2020;2(4):e190064. Link, Google Scholar
38. Cheng JY, Hanneman K, Zhang T, et al. Comprehensive motion-compensated highly accelerated 4D flow MRI with ferumoxytol enhancement for pediatric congenital heart disease. J Magn Reson Imaging 2016;43(6):1355–1368. Crossref, Medline, Google Scholar
39. Koo TK, Li MY. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 2016;15(2):155–163. [Published correction appears in J Chiropr Med 2017;16(4):346.] Crossref, Medline, Google Scholar
40. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33(1):159–174. Crossref, Medline, Google Scholar
41. Cicchetti DV. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychol Assess 1994;6(4):284–290. Crossref, Google Scholar
42. Hoffmann R, von Bardeleben S, ten Cate F, et al. Assessment of systolic left ventricular function: a multi-centre comparison of cineventriculography, cardiac magnetic resonance imaging, unenhanced and contrast-enhanced echocardiography. Eur Heart J 2005;26(6):607–616. Crossref, Medline, Google Scholar
43. Paetsch I, Jahnke C, Ferrari VA, et al. Determination of interobserver variability for identifying inducible left ventricular wall motion abnormalities during dobutamine stress magnetic resonance imaging. Eur Heart J 2006;27(12):1459–1464. Crossref, Medline, Google Scholar
44. Neizel M, Lossnitzer D, Korosoglou G, et al. Strain-encoded MRI for evaluation of left ventricular function and transmurality in acute myocardial infarction. Circ Cardiovasc Imaging 2009;2(2):116–122. Crossref, Medline, Google Scholar
45. Sugimoto T, Dulgheru R, Bernard A, et al. Echocardiographic reference ranges for normal left ventricular 2D strain: results from the EACVI NORRE study. Eur Heart J Cardiovasc Imaging 2017;18(8):833–840. Crossref, Medline, Google Scholar
46. Schuster A, Stahnke VC, Unterberg-Buchwald C, et al. Cardiovascular magnetic resonance feature-tracking assessment of myocardial mechanics: intervendor agreement and considerations regarding reproducibility. Clin Radiol 2015;70(9):989–998. Crossref, Medline, Google Scholar
47. Morton G, Schuster A, Jogiya R, Kutty S, Beerbaum P, Nagel E. Inter-study reproducibility of cardiovascular magnetic resonance myocardial feature tracking. J Cardiovasc Magn Reson 2012;14(1):43. Crossref, Medline, Google Scholar
48. Mangion K, Burke NMM, McComb C, Carrick D, Woodward R, Berry C. Feature-tracking myocardial strain in healthy adults—a magnetic resonance study at 3.0 tesla. Sci Rep 2019;9(1):3239. Crossref, Medline, Google Scholar
49. Morales MA, van den Boomen M, Nguyen C, et al. DeepStrain: a deep learning workflow for the automated characterization of cardiac mechanics. Front Cardiovasc Med 2021;8:730316. Crossref, Medline, Google Scholar
50. Qin C, Wang S, Chen C, Qiu H, Bai W, Rueckert D. Biomechanics-informed neural networks for myocardial motion tracking in MRI. In: Martel AL, Abolmaesumi P, Stoyanov D, et al, eds. Medical Image Computing and Computer Assisted Intervention—MICCAI 2020. Lecture Notes in Computer Science, vol 12263. Springer, 2020; 296–306. Google Scholar
51. Morales MA, Cirillo J, Nakata K, et al. Comparison of DeepStrain and feature tracking for cardiac MRI strain analysis. J Magn Reson Imaging 2022. 10.1002/jmri.28374. Published online July 28, 2022. Medline, Google Scholar
52. Morales MA, Snel GJH, van den Boomen M, et al. DeepStrain evidence of asymptomatic left ventricular diastolic and systolic dysfunction in young adults with cardiac risk factors. Front Cardiovasc Med 2022;9:831080. Crossref, Medline, Google Scholar
53. Aikawa T, Oyama-Manabe N. Editorial for “Comparison of DeepStrain and feature tracking for cardiac MRI strain analysis”. J Magn Reson Imaging 2022. 10.1002/jmri.28394. Published online August 26, 2022. Medline, Google Scholar

Article History

Received: Sept 22 2022
Revision requested: Nov 25 2022
Revision received: Mar 7 2023
Accepted: Mar 20 2023
Published online: May 11 2023

Vol. 5, No. 3

Supplemental Material

Abbreviations

Abbreviations:
AUC	area under the receiver operating characteristic curve
CNN	convolutional neural network
DLSS	deep learning synthetic strain
4D	four-dimensional
ICC	intraclass correlation coefficient
SSFP	steady-state free procession

Metrics

Altmetric Score

PDF download