3D cardiac navigation with rapid multi shot EPI

Summary

To assess a rapid 3D multishot EPI acquisition as an improved cardiac respiratory navigator.

Background

3D EPI navigators are a robust real-time brain navigation tool [1], they allow rapid online reconstruction and image registration (< 80 ms). A thoracic EPI volume can be acquired in 200 ms, thus allowing real-time navigation. An analysis of the EPI navigators’ stability and variance when registering the heart is presented.

Methods

EPI parameters were: flip angle 2°, FOV (v1) 332 x 221 x 144 mm³ or (v2 to v4) 400 x 300 x 150 mm³, acquisition matrix 48 x 36 x 18, TR 14 ms, TE 6.3 ms, slice partial Fourier 6/8, and bandwidth 3858 Hz/pixel, acquisition time 200 ms. The registration region of interest (ROI), the heart, was identified using the adjustment volume. The images were reconstructed in real-time and fed into a modified 3D PACE rigid body registration [2] which registered the ROI to that of the first navigator’s volume.

Four volunteers (mean age 32 +/- 7 years) were scanned on a Siemens 3T. For each, a scan was acquired with 50 navigator volumes, one per R-R interval. Each volunteer held their breath at end expiration for +/- 10 heart beats, then at end inspiration for +/- 10 heart beats, repeating this until the end of the scan. A fifth volunteer was instructed to breathe deeply for the entire scan. Finally the navigators’ impact on Mz was measured with a Bloch simulator [3].

Results

A sample navigator volume and the translations and rotation estimates from one volunteer are shown in the figure. The standard deviation of each motion estimate, calculated as in [4] and by excluding transitions zones, are presented in the table. These measures demonstrate an upper limit on registration variance/stability of 0.6 mm and 0.5°. The motion estimates for the fifth volunteer, with deep breathing, exceeded 4 mm or 4° in all measures. The Bloch simulator shows that the sum effect of the 2° flip angles reduces the Mz by 0.7%.

Figure 1

Example navigator volume and registration result from a volunteer

Table 1 Standard deviation of registration during breathhold periods

Conclusions

EPI proves to be rapid, reliable and consistent as a heart navigator. Its 2° flip angle has a minimal effect on the image contrast (Mz). The real-time nature of this navigator would prove particularly beneficial for techniques like spectroscopy, high resolution imaging, and various forms of functional cardiac imaging.

Funding

This work was supported by the Medical Research Council [G0900883].