Bandpass optical flow for tagged MR imaging

Abstract

MR tagging has shown great promise for detailed noninvasive cardiac motion imaging. We consider here the use of low-frequency tags coupled with gradientbased optical flow estimation to compute cardiac motion. A multiple constraint optical flow method for tagged MRI is formulated by exploiting the Fourier content of the tagged images. The method is validated on simulated tagged data.&words-optical flow, motion estimation, cardiac tagging.

Imaging heart motion using harmonic phase MRI

“…Optical flow methods have also been explored in the analysis of tagged MR image sequences [19]- [21]. In this approach, sinusoidal tag patterns are used instead of saturated planes.…”

“…Generally, regularization is required in order to account for the fact that brightness gradients contain information about motion only in the direction of the gradient. It is possible, however, to combine information from multiple directions, similar in spirit to what is done in planar tagging, in order to produce optical flow estimates without regularization [21]. One limitation of this overall approach is that it is difficult to measure large motions such as that occurring between end diastole and end systole.…”

Recent Advances in Signal Processing

“…Technical details about this type of methods will be explained later. Nevertheless, there have been several strategies to perform the tracking of lines in the myocardium using such techniques (Prince & McVeigh, 1992;Gupta & Prince, 1995;Gupta et al, 1997;Dougherty et al, 1999). The main limitation of these methods is the difficulty to follow large displacements.…”

A non-rigid registration approach for quantifying myocardial contraction in tagged MRI using generalized information measures

“…Measuring myocardial deformations from tagged MR sequences relies on estimating a dense displacement field consistent with the motion of the structured tagging pattern (Kerwin et al, 2001;Amini and Prince, 2001;McVeigh and Ö ztü rk, 2001). Classical approaches comprise: (i) differential optical flow-based methods coupled with intensity correction schemes (Prince and McVeigh, 1992;Denney, 1994;Denney and Prince, 1994;Gupta and Prince, 1995;Gupta et al, 1997;Dougherty et al, 1999); (ii) phase-based optical flow methods such as HARP (Osman et al, 1999;Osman and Prince, 2000;Amini and Prince, 2001); (iii) segmentation-based approaches combining tag segmentation, sparse motion estimation along tags, and dense motion interpolation over the image domain (Kumar and Goldgof, 1994;Guttman et al, 1994;Denney and Prince, 1995;Kraitchman et al, 1995;Young et al, 1995;Park et al, 1996;Declerck, 1997;Guttman et al, 1997;Amini et al, 1998;Denney, 1999;Huang et al, 1999;Kerwin and Prince, 1999;Clarysse et al, 2000;Haber et al, 2000;Ö ztürk and McVeigh, 2000;). However, their clinical applicability is questioned by intrinsic limitations: differential optical flow fails for large displacements due to the local nature of constraint equations; phase-based optical flow is limited to 1-D tags and proves to be sensitive to artifacts; tag segmentation often requires supervision to deal with tag contrast attenuation over time due to T1 demagnetization (tag fading out), and tag destruction along myocardial boundaries induced by blood flows within the cardiac chambers.…”