Diastolic function imaging: a comparison of real-time phase contrast magnetic resonance (CMR) imaging with segmented phase contrast CMR and Doppler echocardiography

Background

CMR measurement of mitral inflow velocities for the assessment of diastolic function is often infeasible in patients with dyspnea - patients who may benefit the most - due to their inability to breath-hold. Although real-time phase contrast (RT-PC) imaging may overcome this limitation, it has not been systematically evaluated. The objective of this study was to assess the accuracy of RT-PC for the measurement of mitral inflow velocities against segmented PC CMR and Doppler echocardiography.

Methods

37 healthy volunteers (aged 28 ± 10 years, 20 males) had echo and CMR studies within a week. Early (E) and late (A) mitral inflow velocities were measured by echo, segmented, and RT-PC CMR (Figure). The E and A velocities were obtained by averaging data from 2 heart beats by RT-PC and 3 heart beats by echo. RT-PC parameters were: TR/TE = 14.0ms/2.3ms, water excitation flip angle=25○,10mm slice, 90 x128 matrix, EPI factor=15, TSENSE rate=3, and VENC=150cm/s. Shared velocity encoding was used to achieve an effective temporal resolution of 28ms, but true temporal resolution was 56ms. Retro-gated segmented PC acquisition parameters: TR/TE = 4.5/1.9ms, 10mm slice, 100 x 192 matrix, TSENSE rate=3, VENC=150cm/s, true temporal resolution 36ms. E and A velocities, and E/A ratios between RT-PC and segmented PC CMR or Doppler echocardiography were compared using paired t-tests. Agreement between the techniques was assessed using concordance correlation coefficients and Bland-Altman analysis.

Figure 1

Mitral inflow velocities in one volunteer by all three techniques: (A) real-time phase contrast imaging (mean E, A, and E/A were 78cm/s, 25cm/s, and 3.1), (B) segmented phase contrast imaging (mean E, A, and E/A were 75cm/s, 27cm/s, and 2.8), and (C) Doppler echocardiography (mean E, A, and E/A were 77cm/s, 26cm/s, and 3.0).

Results

Mean E velocities by echo, segmented, and RT-PC CMR were 75 ± 15 cm/s, 77 ± 12 cm/s, and 73 ± 12cm/s , respectively. The RT-PC measurements were not different from echo (p=0.3), but were less than segmented PC CMR (p=0.04). The A velocities (38 ± 12 cm/s, 38 ± 11 cm/s, 35 ± 12 cm/s, respectively) were not different between RT-PC CMR and echo or segmented CMR (p=0.3 for both). There was also no difference in the E/A ratios (2.2 ± 0.6, 2.2 ± 0.7, and 2.2 ± 0.9, respectively; p =0.6 for both). There was moderate concordance between RT-PC CMR and segmented CMR and Echo for E, A and E/A ratio (Table 1). Although, the bias in measurement between RT-PC CMR and echo or segmented CMR was small, the LOA was wide.

Table 1

Conclusions

We demonstrate for the first time the use of RT-PC imaging to measure mitral E and A velocities. There was modest agreement between RT-PC CMR and echo and segmented PC CMR. Further refinements of the RT-PC sequences are necessary; however, the use of RT-PC imaging provides an opportunity for wider application in patients who have difficulty with breath holding or arrhythmias.

Funding

National Institute of Health (NIH, R01).