Jerome Lamy1, Jie Xiang1, Felicia Seemann2, Ricardo A Gonzales3, Steffen Huber1, Jeremy Steele4, Einar Heiberg5, and Dana C Peters1
1Radiology and Biomedical Imaging, Yale University, New Haven, CT, United States, 2National Institutes of Health, Bethesda, MD, United States, 3Oxford University, Oxford, United Kingdom, 4Internal Medicine, Yale University, New Haven, CT, United States, 5Lund University, Lund, Sweden
Synopsis
Keywords: Flow, Data Acquisition
Tricuspid regurgitant velocity is a crucial
biomarker in identifying pressure overload in the right heart, associated with
diastolic dysfunction and pulmonary hypertension. 2D phase-contrast cannot quantify this flow,
and echocardiography is used clinically. We developed a phase-contrast method
which utilizes deep-learning algorithms to track the valvular slice in a
cardiac phase-dependent manner, which we call 2.5D flow. We studied its performance in nine healthy
subjects and patients with tricuspid regurgitation. RV stroke volumes
correlated better to forward flow volumes by 2.5D flow vs. static 2D
phase-contrast (ICC=0.88 vs. 0.62). 2.5D
flow characterized regurgitation in a patient.
Introduction
Valve diseases are an important cause of
morbidity and mortality (1). Specifically,
tricuspid valve (TV) regurgitation can be detected in 80% of the general
population and considered pathological (moderate or severe) in 15% (2). Tricuspid valvular
regurgitation is often due to elevated
right ventricle (RV) pressure, commonly seen in pulmonary hypertension (PH)
(3) and patients with
diastolic dysfunction, where tricuspid regurgitant velocity is one of 4
criteria used to identify dysfunction (along with LA volume, E/e’ and E/A)
(4). Thus, evaluation of tricuspid regurgitant velocity
is clinically highly important. A recent study (5) of
diastolic dysfunction by MRI, used vorticity duration as a stand-in for
tricuspid regurgitant flow, highlighting the need for its evaluation. According to the current ACC/AHA guidelines, TV
regurgitation is assessed with a comprehensive transthoracic echocardiography
(TTE) imaging with Doppler interrogation (1) of blood velocities. Cardiac MR is considered more accurate for mitral and tricuspid
regurgitant volumes, using indirect evaluation by subtraction of RV (or LV)
stroke volume (SV) from pulmonary artery (PA) or Aortic (Ao) forward flow (6,7). However, direct valve flow evaluation by cardiac
MRI is not feasible due to valvular displacement during the cardiac cycle; even
more so for the highly dynamic (translating and rotating) TV (8). 4D flow methods have had success in tricuspid regurgitant velocity
evaluation, using many minutes of scan time, because retrospective valve
tracking can be employed (9,10).
Prospective valve-tracking methods have been
employed to acquire 2D phase-contrast (PC) with a dynamic slice plane
prescription that changes over the cardiac cycle (11,12). We recently used this
approach, but using modern feature-tracking of the mitral valve (13) to enable rapid and
accurate valve-tracking of the simple
mitral valve translations (14). Even so, to obtain accurate displacements and
valvular velocities (needed to correct flow values) often required expert,
tedious, and time-consuming manual annotations.
More recently, we have developed deep-learning algorithms to fully-automatically
track both the mitral valve (MVnet (15)) and also tricuspid
valve insertion points (TVnet) (16 ,17) with the TV exhibiting greater motion
including rotation vs. the mitral valve.
In this study, we utilize a 2D PC
sequence, with dynamic slice-prescription based on automatic tracking in 2- and
4-chamber RV cines to determine phase-dependent slice translation and rotation,
for prospective valve-tracking PC. This
PC approach is called 2.5D PC, because of the partial 3rd dimension.Methods
Figure 1 shows
the workflow for 2.5D PC. First, RV 2 and 4 chamber cines are acquired and exported
to an offline computer for automated tracking of the valve-insertion points,
using TVnet. This automated tracking produces the center point of the TV plane
and it’s the normal to the TV plane for each time-point in the cardiac cycle.
This is automatically input to the customized MRI sequence via a USB device. During the breath-hold, the slice geometry is
updated by the sequence at each cardiac phase to match the valve position and
orientation.
Nine
healthy volunteers (36±16y, BMI of 24.9±3.8, 4 females) underwent a cardiac MR (3T Siemens, Erlangen, Germany)
that included a standard 4-chamber cine and the less common RV 2-chamber cine,
both used for automated valve-tracking by TVnet. The study was approved by our local IRB and
all subjects provided informed signed consent. The 2D-PC scan protocol for the
TV was: FOV: 380mm, acquisition matrix=256x208, repetition time=5.3ms, echo
time=3.4ms, flip angle=15°, voxel size= 1.48x1.48x5-6 mm3, GRAPPA=2,
partial Fourier 6/8, through-plane flow-encoding with a VENC of 100cm/s to
150cm/s; temporal resolution of 42ms. This
acquisition was performed for a static TV plane coinciding with the valve plane
in late-systole and with a dynamic valve-tracking. Standard planimetry of the cine
stack yielded SVs, and standard PA and Ao PC were performed to compare
resultant SV values. PC analysis was done using Segment software (18), including eddy
current compensation, using cardiac phase-dependent ROIs to identify static
tissue. The flow velocities were corrected for relative motion of the valve(19), on a pixel by pixel basis, for both static and 2.5D PC flow evaluation.Results
Figure 2 shows a
tricuspid flow curve, presenting the flow by PC for the static plane, and the
valve-tracking PC. The valve-tracking plane yields a more physiological curve
in general, with mainly zero flow in systole, when the valve is shut and flow
peaks in diastole corresponding to the E and A wave. 2.5D PC forward flow compared well to stroke
volumes by planimetry (RVSV, ICC=0.88, bias ± 2SDs of -2.5 ± 6.3mls, Figure 3B;
PA flow, ICC=0.72, bias ± 2SDs of -5.1 ± 15.9mls; LVSV and Aortic flow also
agreed well), as expected in healthy subjects. As shown in Figure 3, both
static and 2.5D PC were well correlated to RVSV when corrected for relative
velocities of the valve, but the 2.5D PC showed a slope closer to unity, a
smaller bias, and a much stronger ICC. Figure
4 shows the performance of 2.5D PC in a patient with regurgitation.Discussion
The 2.5D PC method was
validated for forward flow, with performance similar to that reported for 4D
flow techniques (9,10), and it
accurately follows the tricuspid valve. Further studies in patients with
regurgitation are needed to define 2.5D PC’s ability to detect regurgitant
jets.Acknowledgements
NIH R01HL144706.References
1. Otto CM, Nishimura RA, Bonow RO,
Carabello BA, Erwin JP, Gentile F, Jneid H, Krieger EV, Mack M, McLeod C,
O'Gara PT, Rigolin VH, Sundt TM, Thompson A, Toly C. 2020 ACC/AHA Guideline for
the Management of Patients With Valvular Heart Disease: A Report of the
American College of Cardiology/American Heart Association Joint Committee on
Clinical Practice Guidelines. Circulation 2021;143(5):e72-e227.
2. Singh JP, Evans JC, Levy D, Larson
MG, Freed LA, Fuller DL, Lehman B, Benjamin EJ. Prevalence and clinical
determinants of mitral, tricuspid, and aortic regurgitation (the Framingham
Heart Study). The American journal of cardiology 1999;83(6):897-902.
3. Mutlak D, Aronson D, Lessick J,
Reisner SA, Dabbah S, Agmon Y. Functional Tricuspid Regurgitation in Patients
With Pulmonary Hypertension Is Pulmonary Artery Pressure the Only Determinant
of Regurgitation Severity? Chest 2009;135(1):115-121.
4. Nagueh SF, Smiseth OA, Appleton CP,
Byrd BF, Dokainish H, Edvardsen T, Flachskampf FA, Gillebert TC, Klein AL,
Lancellotti P, Marino P, Oh JK, Popescu BA, Waggoner AD, Echocardiography AS,
Imaging EAC. Recommendations for the Evaluation of Left Ventricular Diastolic
Function by Echocardiography: An Update from the American Society of
Echocardiography and the European Association of Cardiovascular Imaging. Eur
Heart J-Card Img 2016;17(12):1321-1360.
5. Ramos JG, Fyrdahl A, Wieslander B,
Thalen S, Reiter G, Reiter U, Jin N, Maret E, Eriksson M, Caidahl K, Sorensson
P, Sigfridsson A, Ugander M. Comprehensive Cardiovascular Magnetic Resonance
Diastolic Dysfunction Grading Shows Very Good Agreement Compared With
Echocardiography. Jacc-Cardiovasc Imag 2020;13(12):2530-2542.
6. Uretsky S, Gillam L, Lang R,
Chaudhry FA, Argulian E, Supariwala A, Gurram S, Jain K, Subero M, Jang JJ,
Cohen R, Wolff SD. Discordance between echocardiography and MRI in the
assessment of mitral regurgitation severity: a prospective multicenter trial. J
Am Coll Cardiol 2015;65(11):1078-1088.
7. Uretsky S, Argulian E, Narula J,
Wolff SD. Use of Cardiac Magnetic Resonance Imaging in Assessing Mitral
Regurgitation: Current Evidence. J Am Coll Cardiol 2018;71(5):547-563.
8. Ton-Nu TT, Levine RA, Handschumacher
MD, Dorer DJ, Yosefy C, Fan DL, Hua LQ, Jiang L, Hung J. Geometric determinants
of functional tricuspid regurgitation - Insights from 3-dimensional
echocardiography. Circulation 2006;114(2):143-149.
9. Feneis JF, Kyubwa E, Atianzar K,
Cheng JY, Alley MT, Vasanawala SS, Demaria AN, Hsiao A. 4D Flow MRI
Quantification of Mitral and Tricuspid Regurgitation: Reproducibility and
Consistency Relative to Conventional MRI. J Magn Reson Imaging
2018;48(4):1147-1158.
10. Driessen MMP, Schings MA, Sieswerda
GT, Doevendans PA, Hulzebos EH, Post MC, Snijder RJ, Westenberg JJM, van Dijk
APJ, Meijboom FJ, Leiner T. Tricuspid flow and regurgitation in congenital
heart disease and pulmonary hypertension: comparison of 4D flow cardiovascular
magnetic resonance and echocardiography. J Cardiovasc Magn Reson 2018;20(1):5.
11. Kozerke S, Schwitter J, Pedersen EM,
Boesiger P. Aortic and mitral regurgitation: Quantification using moving slice
velocity mapping. J Magn Reson Imaging 2001;14(2):106-112.
12. Kozerke S, Scheidegger MB, Pedersen
EM, Boesiger P. Heart motion adapted cine phase-contrast flow measurements
through the aortic valve. Magn Reson Med 1999;42(5):970-978.
13. Seemann F, Heiberg E, Carlsson M,
Gonzales RA, Baldassarre LA, Qiu M, Peters DC. Valvular imaging in the era of
feature-tracking: A slice-following cardiac MR sequence to measure mitral flow.
J Magn Reson Imaging 2019.
14. Seemann F, Pahlm U, Steding-Ehrenborg
K, Ostenfeld E, Erlinge D, Dubois-Rande JL, Jensen SE, Atar D, Arheden H,
Carlsson M, Heiberg E. Time-resolved tracking of the atrioventricular plane
displacement in Cardiovascular Magnetic Resonance (CMR) images. Bmc Med Imaging
2017;17.
15. Gonzales RA, Seemann F, Lamy J,
Mojibian H, Atar D, Erlinge D, Steding-Ehrenborg K, Arheden H, Hu CX, Onofrey
JA, Peters DC, Heiberg E. MVnet: automated time-resolved tracking of the mitral
valve plane in CMR long-axis cine images with residual neural networks: a
multi-center, multi-vendor study. J Cardiovasc Magn R 2021;23(1).
16. Gonzales R, Lamy J, Thomas K, Zhang Q,
Shanmuganathan M, Heiberg E, Ferreira V, Piechnik S, Peters DC. TVnet:
automated global analysis of tricuspid valve plane motion in CMR long-axis
cines with residual neural networks. European Heart Journal-Cardiovascular
Imaging, 23(S2), pp36-37 2022.
17. Gonzales R, Lamy J, Seemann F,
Arvidsson P, Murray V, Heiberg E, Peters D. TVnet: Automated Time-Resolved
Tracking of the Tricuspid Valve Plane in Long-Axis Cine Images with a Dual
Stage Deep Learning Pipeline. In International Conference on Medical Image
Computing and Computer-Assisted Intervention
(MICCAI) (pp 567-576) Springer, Cham 2021.
18. Heiberg E, Sjogren J, Ugander M,
Carlsson M, Engblom H, Arheden H. Design and validation of Segment--freely
available software for cardiovascular image analysis. Bmc Med Imaging
2010;10:1.
19. Kayser
HWM, Stoel BC, vanderWall EE, vanderGeest RJ, deRoos A. MR velocity mapping of
tricuspid flow: Correction for through-plane motion. J Magn Reson Imaging
1997;7(4):669-673.