Aortic Pulse Wave Velocity (PWV) Calculation

This third calculation allows the velocity of a pressure pulse wave (the pulse wave velocity) to be calculated in a phase-contrast angiography image. The propagation delay between two regions of interest is computed by examining the time/velocity profile in each region. You can also use this tool to examine the volume flow rate through the cardiac cycle in the vessels, although if that is the only thing you want to do, then you would be better using the volume flow rate calculation (VFR).

In this example, we will use a single axial slice phase-contrast angiogram through the ascending and descending aorta to calculate the pulse wave velocity (PWV) around the aortic arch. An MRI phase-contrast angiography sequence results in both a magnitude image and a phase image; the phase image is used to calculate the blood flow velocities, but we will use the magnitude image to outline the ascending and descending aorta.

The phase image is shown as a movie below.

(Animated: View cine on-line version)

Phase images have intensity ranges that map to phase changes of +π .. -π. The actual intensity values that map to this phase range can be -4096 .. +4096 (typical of images from Siemens MRI scanners), 0 .. +4096 or -3142 .. +3142 (typical of Philips images). Jim can automatically convert from the image intensity values to phase values.

We will outline the ascending and descending aorta on a few of the magnitude cine frames, propagate the ROIs to all time points and then go on to calculate the PWV.

Having loaded the magnitude image into Jim's main display, we launched the ROI Propagation toolkit and, using Spline ROIs, outlined the ascending and descending aorta at 4 time points: at the beginning and end of the cardiac cycle, and two points around systole. One of these time points is illustrated below:

Phase-contrast angiogram magnitude image, with ascending and descending aorta outlines shown at one of the time-points

Initially, we used the settings:

• Number of shape coefficients = 4. The vessels have a simple outline shape, and 4 shape coefficients is enough to capture the variations in shape.
• Changes are periodic with time. This is an E.C.G. triggered cine sequence, where the last image frame joins back to the first at the end/start of the cardiac cycle.
• The number of time coefficients was not set manually.

However, these settings resulted in a 'bleb' forming on the ascending aorta in some of the image frames, where the image contrast was poor, as illustrated below.

'Bleb' seen on the ROIs in some frames of the phase-contrast angio magnitude image

During the propagation, in some of the image frames, the contrast is not good enough to prevent the ROIs from wandering from the outline of the aorta. This was corrected by manually setting the number of time coefficients to 30. Running the propagation again the resulted in a smoother transition from frame-to-frame, and the outline of the aorta was then followed accurately across all frames.

Calculating the PWV

We will now go on to calculate the PWV of the aorta. Select the 'PWV' tab in the 'Propagated ROI Calculations' panel:

The setup is the same as for the VFR calculation, but with these additional settings:

The propagation distance around the aortic arch. In order to find this distance, you will need to load an oblique sagittal image showing the arch in the main Jim display, and the phase-contrast angiogram in the cross-reference image display. With the cross-reference image location shown, you can now use Open Spline ROIs to draw the inner and outer borders of the lumen of the aortic arch, as illustrated below.

The inner and outer borders of the lumen of the aortic arch drawn as Open Spline ROIs (red lines), in order to calculate the pulse wave propagation distance. The location of the cross-referenced image is indicated by the yellow line. The cross-referenced image is shown below.

The cine magnitude image is loaded into the cross-reference display, to indicate the location of the axial slice on the arch.

The propagation distance was calculated as the average of the lengths of the two ROIs delineating the inner and outer borders of the lumen - this is the approximate length of the mid-line of the aorta.

There are two ways that the time lag between the two locations on the aortic arch can be calculated.

1. Using the cross-correlation between the flow velocity profiles at the two locations. By default, this tool uses the velocity waveform over the whole of the cardiac cycle to assess the pressure wave lag. However, you can change it to use only a certain range of frames of the cine sequence by entering the first and/or the last of frame numbers that you want to use. Generally, it may be better to use only the first half of the cardiac cycle.
2. Using the time-to-foot method. This method determines the lag by estimating the time at which the velocity profile first begins to rise above a baseline level for both the ascending and descending aorta locations. This is done by modelling the upslope to peak velocity as a gamma variate function. You can restrict the modelled range for the gamma variate by setting the first and/or the last of frame numbers.

Click the button to start the area and PWV calculation.

First, the volume flow rate through each ROI is calculated over the cardiac cycle. A graph showing the volume flow rate through each ROI then pops up:

Then a dialog pops up with a report on the calculated pulse wave velocity:

The dialog shows:

• The calculated time lag between the volume flow rate waveforms.
• The calculated pulse wave velocity.
• The maximum flow velocity (averaged over each ROI) through the cardiac cycle.
• The maximum volume flow rate through each ROI through the cardiac cycle.

You can save these data to a report, either in plain text format, or in PDF format. PDF reports can also be sent to a DICOM sever.