Simultaneous acquisition of multiple resolution images for dynamic contrast enhanced imaging of the breast

Abstract

An imaging technique is described that allows the reconstruction of a series of images at high temporal rates, while simultaneously providing images at high spatial resolution. The method allows one to arbitrarily choose from among several combinations of temporal/spatial resolutions during postprocessing. This flexibility is accomplished by strategically interleaving multiple undersampled projection reconstruction datasets (or subapertures), in which each set can be used to reconstruct a high temporal resolution image. Gadolinium contrast-enhanced dynamic imaging is commonly used to diagnose suspicious breast lesions. Both the kinetics of contrast enhancement (1-4) and lesion morphology (5-7) are considered in determining the malignancy of these lesions. The quantitative modeling of enhancement kinetics is best achieved when images are acquired at high temporal rates, on the order of tens of seconds (8,9), while obtaining detailed lesion morphology requires high spatial resolution. Unfortunately, simultaneous acquisition of both high spatial and temporal resolutions is difficult, due to their diverging demands. The high spatial resolution imaging required to distinguish features necessary for the fine architectural information requires a relatively long time to acquire. For example, to image the entire breast a 3D acquisition of 32 slices with a sampling matrix of 512 ϫ 384, takes ϳ100 sec using a repetition time (TR) ϭ 8 ms. Acquiring fewer slices or reducing the matrix size will speed acquisition, but at the cost of coverage or spatial resolution.In order to improve the temporal resolution of the image series, the well-known "keyhole" technique (10,11) has been used in some breast imaging studies (1). In this technique, only the low spatial frequencies along the phase encoding (k y ) direction are acquired at short intervals and the full resolution images are reconstructed by substituting the outer k y -space regions by reference data collected either pre-or postcontrast. One of the drawbacks of this technique is that artifacts can arise from mixing the constantly updated low spatial frequency data with the high frequency data acquired at different (noncontiguous) time periods (9,12,13). Besides potential subject motion between the acquisition of different k-space regions, the high resolution features of the final image may not be enhanced. Other related acquisition schemes have been developed to reduce these artifacts, such as CURE, a pseudo-randomized view acquisition scheme (14), and 3D TRICKS, in which the high spatial frequency data are also periodically updated (although not as frequently as the low spatial frequency) (15,16). Although they may provide more accurate, time-resolved images, the inevitable mixing of old and new data that are noncontiguous in time still occurs with these methods, potentially leading to measurement errors.In this work, a projection reconstruction (PR) imaging technique is described that allows acquisition of high temporal resolution images while simultaneously allowing the r...

Undersampled projection reconstruction for active catheter imaging with adaptable temporal resolution and catheter‐only views

“…In contrast to techniques that interleave projections and then reconstruct each set singly or in combination (18,28), the present method col- lects any Np, on request. The PR trajectory has the characteristic that the temporal resolution can be modified without reduction in nominal spatial resolution (see Figs.…”

Alignment of multimodality, 2D and 3D breast images

“…Pre-and postcontrast (Gadolinium) images are reconstructed using a custom technique. 28,29 Voxel sizes for T1 images were 0.9x0.9x2 to 4 mm 3 with 256x256 images. The pre-and post-contrast images were 0.5x0.5x2 to 4mm 3 and 512x512 pixels, and are illustrated in Figure 2.…”

Reconstruction of undersampled dynamic images by modeling the motion of object elements

“…Then the equations equivalent to Eq. [2] in the k-t space domain for the undersampled case can be written as:…”

“…Clinical applications include cardiac imaging (1), contrastenhanced imaging (2), kinematics of joints and organs (3,4), functional MRI (fMRI) (5), and real-time interventional imaging (6). The simultaneous spatial and temporal resolution desired in these applications is limited due to data-acquisition time constraints.…”