Prospective Motion Correction

Subject motion is a pervasive problem in MR neuroimaging that can affect both functional and anatomical scans. Post-processing methods can retrospectively improve the data quality to a certain degree, but cannot remove all motion-induced artifacts. Prospective motion correction (PMC) is an approach that attempts to reduce motion artifacts at the time of data acquisition by measuring the subject motion in real time and dynamically updating the imaging protocol to track the motion.

 

PMC System

The PMC system uses an optical camera (Kineticor, HI, USA) mounted on the inside of the scanner bore to track the motion of a passive Moire phase marker at 80 Hz frame rate (Maclaren et al., 2012). The Moire phase marker allows the three translational and three rotational degrees of freedom to be measured with precision on the order of tens of microns for the translations and hundredths of degrees for the rotations. Figure 1 shows a subject in a 32-channel head coil with the marker attached via a mold to the upper teeth. The information containing the position and orientation of the marker is sent to the scanner host computer, where the data are used to dynamically update the imaging FOV such that it follows the movement of the marker (Herbst et al., 2014; Herbst et al., 2012; Speck et al., 2006; Zaitsev et al., 2006). Example motion traces measured by the camera are shown in Figure 2.

 

 

fig1

Fig 1: Image of a volunteer with the tracking marker securely attached to the upper teeth using a molded mini-bite bar. The picture is taken from approximately the same angle as where the camera would be located in the scanner bore.

 

fig2

Fig 2: Motion measurements from the PMC camera showing the subject translation (A) and rotation (B).

 

Correction of functional MRI data

The PMC system has been tested for a 3D EPI sequence (Lutti et al., 2013) during non-task fMRI data acquisition (for tSNR evaluation) and during task-fMRI data acquisition (for t-score evaluation). Figure 3 shows the tSNR distributions pooled over 5 healthy subjects, where the inset image of the tSNR map is from one volunteer. The no motion distributions of tSNR were very similar between PMC on and PMC off, and the runs with PMC on had clearly higher overall tSNR values for comparable levels of motion (43% improvement).

The results from two task fMRI experiments are shown in Figure 4. Representative t-score maps of activated voxels (uncorrected p < 0.001) are overlaid on a high resolution structural image for each of the four conditions of the 2x2 factorial design for both motor and visual tasks.

 

fig3

Fig 3: Distribution of tSNR values pooled over 5 subjects from a 3D EPI sequence at 1.5 mm isotropic resolution. The 2x2 design compares PMC On vs PMC Off and motion vs no motion.

 

fig4

Fig 4: T-scores from block-designed fMRI tasks for the same 3D EPI sequence.

 

Correction of anatomical MRI data

Anatomical data sets are typically acquired over a period of minutes making these acquisitions susceptible to volunteer or patient motion. If quantitative data is to be produced, it is typical to combine multiple data sets in which a parameter, e.g. flip angle, has been altered to interrogate the quantitative parameter of interest, e.g. longitudinal recovery rate, R1 (=1/T1 time). This further increases the likelihood of motion artefact corrupting the data. At the WTCN, we use the multiple parameter mapping (MPM) approach in which we combine multiple data sets to produce four different quantitative maps. Particular emphasis is placed on short scan times, high precision and minimal bias in the parameter estimates. Short scan times are not only efficient but also reduce sensitivity to motion. A complimentary approach is to use prospective motion correction (PMC). This method utilises a camera placed in the bore of the magnet that tracks volunteer movement and updates the imaging system in real time. Below is an example of a quantitative map of R1 values created from data acquired with the volunteer moved throughout the acquisition

 

fig5

Fig 5: When PMC is not used the maps have artefacts (left column), even in the no motion case, and structures are not well resolved. When PMC is used to track the movement image quality is greatly improved (right column) and structure definition improves, e.g. the boundary of the corpus callosum, the cerebellum as well as individual sulci and gyri.

 

 

Primary contact

Nikolaus Weiskopf (n.weiskopf «at» ucl.ac.uk)

 

References

Herbst, M., Maclaren, J., Lovell-Smith, C., Sostheim, R., Egger, K., Harloff, A., Korvink, J., Hennig, J., Zaitsev, M., 2014. Reproduction of motion artifacts for performance analysis of prospective motion correction in MRI. Magn Reson Med 71, 182-190.

Herbst, M., Maclaren, J., Weigel, M., Korvink, J., Hennig, J., Zaitsev, M., 2012. Prospective motion correction with continuous gradient updates in diffusion weighted imaging. Magn Reson Med 67, 326-338.

Lutti, A., Thomas, D.L., Hutton, C., Weiskopf, N., 2013. High-resolution functional MRI at 3 T: 3D/2D echo-planar imaging with optimized physiological noise correction. Magn Reson Med 69, 1657-1664.

Maclaren, J., Armstrong, B.S., Barrows, R.T., Danishad, K.A., Ernst, T., Foster, C.L., Gumus, K., Herbst, M., Kadashevich, I.Y., Kusik, T.P., Li, Q., Lovell-Smith, C., Prieto, T., Schulze, P., Speck, O., Stucht, D., Zaitsev, M., 2012. Measurement and correction of microscopic head motion during magnetic resonance imaging of the brain. PLoS One 7, e48088.

Maclaren, J., Herbst, M., Speck, O., Zaitsev, M., 2013. Prospective motion correction in brain imaging: a review. Magn Reson Med 69, 621-636.

Speck, O., Hennig, J., Zaitsev, M., 2006. Prospective real-time slice-by-slice motion correction for fMRI in freely moving subjects. MAGMA 19, 55-61.

Zaitsev, M., Dold, C., Sakas, G., Hennig, J., Speck, O., 2006. Magnetic resonance imaging of freely moving objects: prospective real-time motion correction using an external optical motion tracking system. Neuroimage 31, 1038-1050.

back to physics pages