Rapid flow assessment of congenital heart disease using high spatio-temporal gated spiral phase-contrast MR

Purpose

Many sick adults and children are unable to perform long breath-holds required for conventional, Cartesian phase-contrast (PC) sequences. Using a prospectively-triggered spiral PCMR sequence accelerated with sensitivity encoding (SENSE), it should be possible to achieve high resolution PCMR data in a short breath-hold.

The aim of this study was to compare flow volumes measured using: a) reference free-breathing, gated Cartesian PCMR, b) standard breath-hold, gated, Cartesian PCMR, and c)  gated, spiral, SENSE, breath-hold PCMR.

Methods

40 consecutive children and adults were enrolled in this study (M:22, F:18, age:21.4±13.8 years). Flow was measured in the:

• Ascending aorta (AAO, N=40)

• Main pulmonary artery (MPA, N=38)

• Right pulmonary artery (RPA, N=22)

• Left pulmonary artery (LPA, N=24)

Flow assessment was performed in each vessel using the three sequences above (parameters shown in Table 1).

Table 1

Stroke volume and regurgitation fraction were calculated for each patient. Additionally, Qp/Qs (N=38) and RPA/LPA (N=20) ratios were quantified where possible.

Results

Average scan time was 91±17 seconds for the reference free-breathing sequence, 16±3 seconds for the standard breath-hold sequence, and 5±1 seconds for the spiral breath-hold sequence.

Combining all vessels (N=124), there were no statistical differences in mean stroke volume calculated from the reference free-breathing sequence (60.3±27.3 mL), the standard breath-hold sequence (59.8±27.6 mL) and the spiral breath-hold sequence (59.5±27.1 mL). Bland-Altman analyses are shown in Figure 1. There was no clinically significant bias using either breath-hold sequence (spiral breath-hold: -0.7 mL, standard breath-hold: -0.5 mL). However, the limits of agreement were smaller and the correlation better for the spiral breath-hold compared to the standard breath-hold sequence (-4.4 to 2.9 mL vs. -10.3 to 9.3 mL, respectively).

Figure 1

There was also an excellent agreement in QP/QS and RPA/LPA ratios between all sequences, however the spiral breath-hold sequence was found to be superior to the standard breath-hold sequence in terms of limits of agreement and correlation. There was a small but statistically significant underestimation of regurgitation fraction using the spiral sequence.

Conclusion

Flow volumes can be accurately and reliably quantified using a spiral SENSE PCMR sequence, with high spatio-temporal resolution in a short breath-hold. As the standard method of measuring flow in congenital heart disease is free-breathing, cardiac gated PCMR, this spiral sequence could reduce total flow imaging from ~10 minutes, to <1 minute. This is a marked reduction in total scan time and has implications for patient throughput and compliance for congenital cardiac MR scanning.