Improved motion-correction for MRI with markerless face-tracking

Marchetto, Elisa 2023. Improved motion-correction for MRI with markerless face-tracking. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF (PhD thesis) - Published Version Download (14MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (197kB)

Abstract

Motion artifacts are a well-known problem in MRI. They can extensively reduce image sharpness and resolution, as well as obscure pathologic conditions, which will make the images not suitable for clinical or research purposes. Over the years, multiple motion correction methods have been proposed to compensate for motion artifacts in different MRI applications. In this thesis, we investigate methods to maximize the image quality of brain MR images at different motion regimes, with the goal of obtaining high-quality images in the case of large and continuous motion profiles as might be expected in some children or patients with movement disorders. We describe a new autofocusing algorithm to correct for in-plane translations and rotations without any previous information coming from motion tracking sources. Preliminary results show good motion compensation for 2D translations. However, we show how rotations cannot be accurately estimated at the present stage, which should be investigated in future studies. We analyse the extent of the motion parameters estimation accuracy of a navigator-based motion correction method using simulated data. The navigator relies on GRAPPA reconstruction of the highly accelerated navigator fat-volumes to estimate the motion parameters. Our results suggest that the fat-navigator is capable of compensating for large range of motion, as well as for fast and slow changes in the head position. Better correction is expected if GRAPPA weights are updated throughout the entire duration of the scan. The fat-navigator is then compared with another tracking technique based on structured light to track the subject’s head movements. We present the results obtained from different motion types as well as a method to improve the motion estimation accuracy of the navigator-based technique in the presence of extensive pitch-wise motion using a skull masking approach. Finally, we introduce a method to quickly develop and test motion-robust pulse sequences using an open-source framework to acquire MR images producing low acoustic noise levels, which make them suitable for paediatric/infant age group, where research scans are typically conducted while the subject is sleeping.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Engineering
Uncontrolled Keywords:	1). Motion-Correction 2). Brain MRI 3). Markerless 4). Fat-navigator 5). Face-tracking 6). FatNavs
Date of First Compliant Deposit:	20 April 2023
Last Modified:	21 Apr 2023 11:59
URI:	https://orca.cardiff.ac.uk/id/eprint/158966

Actions (repository staff only)

Edit Item