Retrospective 3D motion correction using spherical navigator echoes

Abstract

Purpose

To develop and evaluate a rapid spherical navigator echo (SNAV) motion correction technique, then apply it for retrospective correction of brain images.

Methods

The pre-rotated, template matching SNAV method (preRot-SNAV) was developed in combination with a novel hybrid baseline strategy, which includes acquired and interpolated templates. Specifically, the SNAV templates are only rotated around X- and Y-axis; for each rotated SNAV, simulated baseline templates that mimic object rotation about the Z-axis were interpolated.

The new method was first evaluated with phantom experiments. Then, a customized SNAV-interleaved gradient echo sequence was used to image three volunteers performing directed head motion. The SNAV motion measurements were used to retrospectively correct the brain images. Experiments were performed using a 3.0 T whole-body MRI scanner and both single and 8-channel head coils.

Results

Phantom rotations and translations measured using the hybrid baselines agreed to within 0.9° and 1 mm compared to those measured with the original preRot-SNAV method. Retrospective motion correction of in vivo images using the hybrid preRot-SNAV effectively corrected for head rotation up to 4° and 4 mm.

Conclusions

The presented hybrid approach enables the acquisition of pre-rotated baseline templates in as little as 2.5 s, and results in accurate measurement of rotations and translations. Retrospective 3D motion correction successfully reduced motion artifacts in vivo.

Introduction

Despite the advent of continually faster acquisition strategies, subject motion remains a problem in MRI and numerous techniques have been proposed to correct it. One such method is the use of navigator echoes [1], first proposed by Ehman et al. in 1989 to measure and correct rigid body motion. Pencil beam navigators, frequently used in cardiac imaging [2], [3] and orbital navigators [4], [5], [6] are limited to motion measurement in one and two dimensions respectively. Three-dimensional (3D) motion can be measured using cloverleaf [7] or spherical [8], [9] k-space navigators. The spherical navigator echo (SNAV), which samples a spherical shell in k-space, can measure motion in 6 degrees of freedom [9] by using relationships that follow from the Fourier shift and rotation theorems [4]. First described by Welch et al. [9], [10] SNAVs promised simultaneous 3D motion measurements but were impractical due to long measurement and processing times. The polar SNAV approach [11] proposed a faster registration technique but translation estimates remained too slow for prospective correction.

These limitations with navigator echoes have led to the development of alternative motion correction strategies – primarily image based and optical-tracking based methods. Image based tracking methods [12], [13], [14], [15], [16], [17], [18], [19] and fat navigators [20], [21], [22] involve the acquisition of low-resolution images to track – and correct for – motion throughout image acquisition. While highly effective, these methods have been applied only to spin-echo based imaging because of the lengthy acquisition and processing of low-resolution images used to quantify motion. Optical tracking methods, which have been used in many studies [23], [24], [25], [26], [27], [28], use external camera systems to measure and correct for head motion in real time. However, these techniques rely on tracking devices mounted on the subject as well as MR-compatible external hardware that require cross-calibration between the tracking system and the scanner's frame of reference.

Therefore, while recent work in the field has largely moved away from k-space navigators, SNAVs remain a promising alternative, as they are very rapid to acquire and have been limited mainly by the processing method. A solution to the slow processing speed was proposed by Liu and Drangova [29], who demonstrated that SNAV processing can be reduced from several seconds to less than 20 ms, using a template matching instead of the iterative registration and minimization algorithm used previously. This technique – referred to as the preRot-SNAV technique – relies on the acquisition of a set of baseline template SNAVs with known rotated trajectories covering the entire range of anticipated motion. While accurate motion tracking was demonstrated with the preRot-SNAV technique, its clinical applicability was limited because the acquisition of the pre-rotated baseline required 26 s of “no motion” during baseline acquisition.

Building on the proposed preRot-SNAV technique, this manuscript presents the first application of SNAVs for in vivo motion correction. Retrospective motion correction of 3D gradient echo images was achieved by implementing a modification of the preRot technique, which acquires a subset of pre-rotated baseline templates followed by offline simulation of axial rotation to each of the acquired SNAV templates. This “hybrid baseline” preRot-SNAV technique is advantageous as it reduces the likelihood of subject motion during template acquisition; it is first evaluated in phantom studies then retrospective motion correction is demonstrated in multiple volunteers with single and multi-channel RF coils.