Personal tools

You are here: Home / Research / Biomedical Image Analysis / Professor Michael Chappell / Vessel Encoded ASL-MRI

Vessel Encoded ASL-MRI

Vessel Encoding extends the non-invasive Arterial Spin Labelling Technique so that different vascular territories or downstream arteries can be identified.

Images from VE-ASL research

Collaborators

Thomas Okell, Peter Jezzard (FMRIB Centre)
Nolan Hartkamp (Utrecht)
Matthias Van Osch (Leiden)

Description

Vessel Encoded ASL (VE-ASL) is a form of selective ASL whereby the flow contributions from individual arteries feeding the brain can be uniquely labelled and thus identified. This might take the form of images of perfusion in the brain with different vascular territories (regions fed by a specific artery) marked or angiographic images of the vasculature showing the source of blood in specific artery segments. VE-ASL works by modulating the ASL label procedure so that during each acquisition a different combinations of arteries are labelled, so that over a series of acquisitions each one makes a unique contribution to the imaging data. The individual contributions can then be separated afterward during analysis.

Inference

We have developed a general framework (essentially a model) that describes how the locations of the arteries and the modulation profile of the ASL label contribute to the overall signal. By taking this mode and using Bayesian inference techniques we are able to extract the individual artery contributions to blood flow or perfusion even in regions where more than one artery supplies blood. The technique requires minimal information from the user beside details of the acquisition sequence, being able to determine the locations of the arteries directly from the data, and can even produce results where the encoding process was sub optimal.

Quantification

VE-ASL that captures dynamic information about the label can be used in combination with kinetic models to quantify perfusion in tissue or blood volume and even flow rate through arteries. This work has built upon the perfusion quantification methods employed by the group in standard ASL perfusion quantification

Software

The inference approach for the analysis of VE-ASL data exists as MATLAB code and is in use by a number of collaborating groups. If you would like to use it in your research then please contact us.

Selected Publications

Chappell, M. A., Okell, T. W., Payne, S. J., Jezzard, P., & Woolrich, M. W.
A fast analysis method for non-invasive imaging of blood flow in individual cerebral arteries using vessel-encoded arterial spin labelling angiography.
Medical Image Analysis. doi:10.1016/j.media.2011.12.004

Okell, T. W., Chappell, M. A., Schulz, U. G., & Jezzard, P.
A kinetic model for vessel-encoded dynamic angiography with arterial spin labeling.
Magnetic resonance in medicine, 68(3), 969–979. doi:10.1002/mrm.23311

Chappell, M. A., Okell, T. W., Jezzard, P., & Woolrich, M. W.
A General Framework for the Analysis of Vessel Encoded Arterial Spin Labeling for Vascular Territory Mapping.
Magnetic resonance in medicine, 64(5), 1529–1539. doi:10.1002/mrm.22524