Prospective motion correction for 2D slice-selective FISP-MRF in the brain using an in-bore camera system
Gregor Körzdörfer1,2, Mario Bacher1,3, Thomas Kluge1, Randall Kroeker1, Dominik Paul1, Josef Pfeuffer1, Bernhard Hensel2, and Mathias Nittka1

1Siemens Healthcare GmbH, Erlangen, Germany, 2Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany, 3CHUV, Centre d'Imagerie BioMédicale, Lausanne, Switzerland


In contrast to motion artifacts in conventional MRI, which can often be identified by visual inspection, the effect can be more subtle in quantitative MRI (qMRI) methods such as Magnetic Resonance Fingerprinting (MRF). Subject motion during qMRI scans can lead to altered parameter maps without affecting their morphologic appearance which limits the user’s possibility to assess the scan quality. One way to mitigate motion artifacts is to track the subject’s movement and prospectively correct for the motion. Here, we present results of applying prospective motion correction using an in-bore camera system for MRF.


Magnetic Resonance Fingerprinting (MRF)1 is a novel technique that yields multi-parametric, quantitative MRI. A series of excitation events generates a signal response which is compared to a set of simulated signals to determine the parameters of the tissue where the signal originates from.

A vast number of parameters affect the actual MR signal, but signal simulations are usually limited to incorporating only the most impacting parameters. Commonly, these are tissue parameters like T1 and T2 relaxation times and magnetic field properties like B12,3 and B01. In MR, an inherent confounding factor is subject motion. Methods for limiting and controlling it are limited, which poses a general problem for MRI because of the long acquisition times compared to other medical imaging modalities.

In MRF, it is often assumed that signals corrupted by motion do not significantly affect the pattern match, since the corrupted signal parts do not have a counterpart in the dictionary1. Motion in slice-selective sequences can be categorized into in-plane and through-plane motion. In-plane motion is supposed to lead to only small changes. The measured signal in one voxel originates from varying spatial locations which can be corrected for by registering the images4,5, but the actual spin evolution is not altered. Through-plane motion leads to a changed spin evolution, because out-of-slice spins that have not been excited before may enter into the slice and contribute an altered signal due to their different excitation history. Several works have shown that MRF is sensitive to such through-plane motion6.7.

In this study, a method is presented to prospectively correct for subject motion during a 2D slice selective FISP MRF8 experiment using an in-bore camera system (KinetiCor Inc., Honolulu, HI, USA).


We used a prototype implementation of the MRF method8 that is based on a steady-state free-precession (SSFP) sequence with a prescan-based9 B1+ correction. A spiral sampling scheme was applied (undersampling factor 48, field of view 300 mm, resolution 1.2 mm, slice thickness 5mm) with a spiral angle increment of 82.5° from shot to shot10,11. A volunteer was scanned on a 3T whole-body scanner (MAGNETOM Vida, Siemens Healthcare, Erlangen, Germany) using this implementation while the motion was recorded using the camera system that tracks a marker placed on the volunteers nose. With active prospective motion correction, RF pulses and gradients of the MRF sequence were updated in real-time throughout the scan to continuously image the same anatomical slice. The scan time per slice was 20 seconds.

The volunteer was advised to perform typical movements during the MRF experiment, nodding and shaking the head. Additionally, a scan without motion was performed. For each motion type, two axial slices were acquired both with and without prospective motion correction. In both cases the detected motion (translation and rotation) was recorded. The resulting T1 and T2 parameter maps were segmented into white matter, grey matter, and cerebrospinal fluid (CSF). Parameter values in the segmented regions were compared in box plots.


Figure 1 displays the recorded motion during the MRF experiments. In the nodding experiment, large translation in y-direction was observed. During the scan with the subject shaking his head, rotation around x- and y-axes dominated the motion pattern.

Figure 2 displays the parameter maps from the motion-corrected and non-motion-corrected MRF experiments as well as the differences of the maps to the scan without motion. T1 values are almost unaffected. In the uncorrected MRF experiments, the deviations tend to be stronger the farther away from the axes of motion the voxel is. The T2 map of the uncorrected shaking-head scan appears blurry by visual inspection.

Figure 3 displays the parameter values in the segmented brain regions as boxplots. T1 distributions are similar throughout the scans, while the motion affected T2 values without prospective motion correction exhibit smaller median values.


MRF results that are corrupted by substantial subject motion can be prospectively corrected by using an in-bore camera system with real-time feedback. Similarly to previous work we found that through-plane motion (e.g. nodding) has a higher effect on MRF results than in-plane motion (e.g. shaking head). T1 values were not altered by motion in our study while a significant effect was found for T2 values. CSF T2 values exhibit high intra-scan as well as inter-scan variation, likely due to flow artifacts.

This work was performed in healthy volunteers, and the possibility to transfer the results to real patients and potentially different motion patterns will be investigated in future work.


We have shown how subject motion in MRF experiments can be corrected by using an in-bore camera system featuring real-time feedback to the scan control software.


No acknowledgement found.


1 Ma D. et al, Magnetic resonance fingerprinting. Nature 2013;

2 Buonincontri G et al, MR fingerprinting with simultaneous B1 estimation. MRM 2016;

3 Cloos MA. et al, Multiparametric imaging with heterogeneous radiofrequency fields. Nat. Commun. 2016;

4 Mehta B. et al, Motion Insensitive Magnetic Resonance Fingerprinting (MORF), ISMRM Workshop on Magnetic Resonance Fingerprinting 2017;

5 Cruz G. et al, Rigid motion‐corrected magnetic resonance fingerprinting. MRM 2018;

6 Yu Z. et al, Exploring the sensitivity of magnetic resonance fingerprinting to motion., MRI. 2018;

7 Körzdörfer G. et al, Evaluating the influence of motion on FISP MRF. ISMRM 2018;

8 Jiang Y. et al, MR fingerprinting using fast imaging with steady state precession (FISP) with spiral readout. MRM 2015;

9 Chung S. et al., Rapid B1+ mapping using a preconditioning RF pulse with TurboFLASH readout, MRM 2010;

10 Pfeuffer, J. et al, Mitigation of Spiral Undersampling Artifacts in Magnetic Resonance Fingerprinting (MRF) by Adapted Interleaf Reordering. ISMRM 2017;

11 Körzdörfer, G. et al, Spatial biases in Magnetic Resonance Fingerprinting parameter maps arising from undersampling patterns. ISMRM 2017;


Figure 1: Recorded motion (translation and rotation) during the MRF experiments. a) without motion, b) nodding, and c) shaking head.

Figure 2: Resulting MRF parameter maps from the experiments with different motion patterns. a) shows T1 maps and c) the corresponding difference maps to the scan without motion. b) displays the T2 maps and d) the difference maps to the scan without motion.

Figure 3: Boxplots displaying the distribution of T1 and T2 values in three segmented brain regions (white matter, grey matter, and cerebrospinal fluid).

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)