Blood Flow Imaging with Ultrafast Doppler

The quantitative evaluation of blood flow by ultrasound is an extremely useful parameter in medicine for the assessment of many organs. Ultrafast Doppler ultrasound provides special temporal coherence and facilitates an increased sensitivity to measuring small blood flow velocities. Having access to microperfusion opens the door to more precise understanding of the perfusion of tumors or organs, such as the brain or the heart.

To set up the Doppler Phantom, first use plastic tubes to connect the peristaltic pump, the blood mimicking fluid reservoir, the pulse dampener and the Doppler flow Phantom. Select the canal with a four millimeter diameter and program the pump to eject 720 milliliters per minute of fluid for 0.3 seconds followed by 0.7 seconds of ejection at 50 milliliters per minute to mimic the systolic and diastolic cardiac phases, respectively. Then run the pump and gently shake the pipes to expel any potential air bubbles.

To set up the ultrafast ultrasound scanner, use the PCI Express link to connect the ultrafast enabled research scanner to the host computer and change the transducer adapter on the scanner to match the probe connector. Connect the probe and run MATLAB to activate the ultrasound scanner license. To set up the ultrasound sequence program, set the imaging depth to 50 millimeters and the focal depth to 35 millimeters.

To design an ultrafast ultrasound sequence, set the imaging depth to 50 millimeters, program three tilted plane waves at minus three, zero and three degrees and set the pulse repetition frequency to 12 kilohertz. Then select four half cycles for the ultrasound wave form, with the center frequency set according to the probe used, and set the total duration to one second. When the sequence parameters have been set, apply ultrasound gel to the probe lens and place the probe onto the Phantom.

Launch the B-mode ultrasound sequence and locate the canal of interest. The fluid will appear darker than the surrounding tissue. Switch to the longitudinal view.

Without changing the position of the probe, end the B-mode sequence and launch the ultrafast sequence acquisition script. When the sequence is over, save the raw data and use the ultrasound system default software to launch the imagery construction script. For clutter filtering, use the MATLAB script to reshape the 3D IQ matrix into a 2D Casorati matrix.

Then use the formula to compute the singular value decomposition of the matrix, and use the spatial singular vectors to compute the spatial similarity matrix C and to identify the blood subspace boundaries. Use the blood space boundary cutoff to filter the IQ data. To compute the power Doppler map, use the formula to integrate the envelope of the filter data along the temporal dimension and display the power Doppler map in a logarithmic scale.

Define a circular region of interest containing one to 30 pixels in the image, and average the filter data signal over the pixels within the region of interest to obtain a vector for the filter data of the relevant number of experimental time points. To compute and display the Doppler spectrogram of these data, set the short time Fourier transform window to a 60 samples Hann window, and set the short time Fourier transform overlap to 90%of the window length. Then, overlay the center frequency at each time point of the spectrogram and use the Doppler formula to convert the frequency values into blood axial velocities.

If a good quality spectrogram has been acquired, the blood velocities can be extracted from any region of interest within the image. In this image of a neonate brain, acquisition vessels with very different flow characteristics, from small cortical venules and arterials to the major pericallosal artery, can be observed. Here, the ability of ultra fast Doppler to extract a blood flow signal in a strongly moving organ such as the myocardium is shown.

Although we demonstrated the use of singular value decomposition, other types of filter can also be used. The ultrafast Doppler allows us to better understand the microperfusion. It's a real revolution in medical imaging.