Test Platform for Developing New Optical Position Tracking Technology towards Improved Head Motion Correction in Magnetic Resonance Imaging.

Publisher: MDPI Country of Publication: Switzerland NLM ID: 101204366 Publication Model: Electronic Cited Medium: Internet ISSN: 1424-8220 (Electronic) Linking ISSN: 14248220 NLM ISO Abbreviation: Sensors (Basel) Subsets: MEDLINE

Original Publication: Basel, Switzerland : MDPI, c2000-

Magnetic Resonance Imaging*/methods ; Head*/diagnostic imaging; Humans ; Head Movements ; Neural Networks, Computer ; Fiducial Markers ; Calibration ; Image Processing, Computer-Assisted/methods ; Deep Learning ; Brain/diagnostic imaging ; Artifacts

Optical tracking of head pose via fiducial markers has been proven to enable effective correction of motion artifacts in the brain during magnetic resonance imaging but remains difficult to implement in the clinic due to lengthy calibration and set up times. Advances in deep learning for markerless head pose estimation have yet to be applied to this problem because of the sub-millimetre spatial resolution required for motion correction. In the present work, two optical tracking systems are described for the development and training of a neural network: one marker-based system (a testing platform for measuring ground truth head pose) with high tracking fidelity to act as the training labels, and one markerless deep-learning-based system using images of the markerless head as input to the network. The markerless system has the potential to overcome issues of marker occlusion, insufficient rigid attachment of the marker, lengthy calibration times, and unequal performance across degrees of freedom (DOF), all of which hamper the adoption of marker-based solutions in the clinic. Detail is provided on the development of a custom moiré-enhanced fiducial marker for use as ground truth and on the calibration procedure for both optical tracking systems. Additionally, the development of a synthetic head pose dataset is described for the proof of concept and initial pre-training of a simple convolutional neural network. Results indicate that the ground truth system has been sufficiently calibrated and can track head pose with an error of <1 mm and <1°. Tracking data of a healthy, adult participant are shown. Pre-training results show that the average root-mean-squared error across the 6 DOF is 0.13 and 0.36 (mm or degrees) on a head model included and excluded from the training dataset, respectively. Overall, this work indicates excellent feasibility of the deep-learning-based approach and will enable future work in training and testing on a real dataset in the MRI environment.

Sci Rep. 2020 Mar 16;10(1):4786. (PMID: 32179823)
Sci Data. 2017 Mar 14;4:170032. (PMID: 28291265)
Nature. 2020 Sep;585(7825):357-362. (PMID: 32939066)
Nat Methods. 2020 Mar;17(3):261-272. (PMID: 32015543)
Magn Reson Med. 2019 Oct;82(4):1527-1540. (PMID: 31081955)
Magn Reson Med. 2010 Jan;63(1):162-70. (PMID: 19918892)
Magn Reson Med. 2019 Jul;82(1):126-144. (PMID: 30821010)
Neuroimage. 2017 Jul 1;154:33-42. (PMID: 27845256)
IEEE Trans Med Imaging. 2019 Feb;38(2):470-481. (PMID: 30138909)
Neuroimage. 2001 Aug;14(2):284-97. (PMID: 11467903)
Magn Reson Med. 2009 Oct;62(4):924-34. (PMID: 19526503)
Magn Reson Med. 1991 Mar;18(1):207-13. (PMID: 2062232)
J Magn Reson Imaging. 2015 Oct;42(4):887-901. (PMID: 25630632)
Magn Reson Med. 2019 Oct;82(4):1452-1461. (PMID: 31045278)
Neuroimage. 2021 Apr 15;230:117756. (PMID: 33460797)
Magn Reson Med. 2019 Sep;82(3):901-910. (PMID: 31006909)
Magn Reson Med. 2012 May;67(5):1237-51. (PMID: 21826729)
Neuroimage. 2006 Jul 1;31(3):1038-50. (PMID: 16600642)
PLoS One. 2012;7(11):e48088. (PMID: 23144848)
Magn Reson Med. 2023 Oct;90(4):1297-1315. (PMID: 37183791)
Phys Med Biol. 2016 Mar 7;61(5):R32-56. (PMID: 26864183)
Magn Reson Med. 2018 Apr;79(4):1911-1921. (PMID: 28722314)
PLoS One. 2015 Jul 30;10(7):e0133921. (PMID: 26226146)
Med Image Anal. 2011 Oct;15(5):708-19. (PMID: 21708477)
Front Neurosci. 2019 Aug 13;13:825. (PMID: 31456656)

RGPIN-2023-04408 Natural Sciences and Engineering Research Council; 30729 Canadian Foundation for Innovation

Keywords: CNN; deep learning; head pose estimation; magnetic resonance imaging; moiré pattern; motion artifacts; motion correction; neuroimaging; optical tracking; synthetic dataset

Date Created: 20240627 Date Completed: 20240627 Latest Revision: 20240629

20240629

PMC11207598

10.3390/s24123737

38931521