Objectives Time-resolved data acquisition with sliding-window conjugate-gradient highly constrained back projection (SW-CG-HYPR) has been proposed to acquire 2D myocardial perfusion images. This method allows increased spatial coverage, better spatial resolution, and improved signal-to-noise ratio (SNR). To further increase the spatial coverage and contrast-to-noise ratio (CNR), we developed a 3D sequence combined with inversion recovery (IR) pre-pulse and SW-CG-HYPR.
Methods Five healthy volunteers were scanned using a 1.5T system (Espree, Siemens, Erlangen, Germany). An ECG-triggered, 3D turbo-FLASH sequence with radial k-space sampling and inversion recovery preparation was used in this study. Within each cardiac cycle, 6 partitions were acquired after a trigger delay time and inversion recovery preparation. Each partition was acquired in a segmented interleaved fashion with 16 projections per heartbeat, and the “composite images” were reconstructed by a sliding window method using k-space data from 10 consecutive cardiac cycles. CG-HYPR method was used to reconstruct the time-resolved images. SW-CG-HYPR allows a 3D acquisition window of 250 ms in each heartbeat. To compare the image quality and verify the signal changes after contrast administration, a conventional IR Turbo-FLASH scan was performed with the same contrast injection scheme.
Results Left ventricle and myocardium signal changes in SW-CG-HYPR images were closely related to those observed in images obtained using the conventional protocol. The mean correlation coefficients between 3D sliding CG-HYPR and 2D single-slice reference images are 0.97, 0.95 for blood and myocardial signals, respectively. With SW-CG-HYPR, 6 partitions were acquired in each cardiac cycle for IR prepared myocardial perfusion imaging, while the conventional protocol only allows 1 slice with IR and 3 slices with saturation recovery preparation. Within the interpolated 12 partitions, the 10 central partitions have no slice aliasing.
Conclusions 3D imaging improves SNR and allows inversion recovery preparation, which improves image contrast over saturation recovery preparation required for multi-slice 2D imaging.