||Transverse blood flow estimation and effective coded excitation using ultrasound
||Jensen, Jørgen Arendt (Biomedical Engineering, Department of Electrical Engineering, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
||Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
||This thesis contains two independent investigation in the area of medical ultrasound. The primary focus
is on a new blood flow velocity estimator. Secondarily coded excitation is investigated.
A technique for estimating the full flow velocity vector is investigated. Unlike the conventional estimators,
that only detect the axial component of the flow, this new method is capable of estimating the
transverse velocity component. The method uses focusing along the flow direction to produce signals
that are influenced by the shift in the position of the scatterer. The signals are then cross-correlated to
find the shift in position and thereby the velocity.
The performance of the method is in this thesis investigated using both simulations and measurements.
The measurements were made with an automatic stepping system, a flow phantom build during the
project period, and in-vivo. Data was acquired with our experimental ultrasound scanner.
Simulations are used to verify the basic function and implementation correctness of the estimator while
the emphasis is on the measurements. In the automatic stepping system an emulated plug flow is measured
under very precise and controllable conditions. This measurement verifies the geometry and basic
parameter setup. The flow phantom is capable of producing a parabolic flow profile with a blood mimicking
fluid. These measurements show how the method behaves under realistic conditions. Measurements
on the abdominal aorta are also presented.
In medical ultrasound improvements of approximately 15-20 dB can be achieved by using coded waveforms.
Exciting the transducer with an encoded waveform necessitates compression of the response,
which can be done either before beamforming (pre-compression) or after beamforming (post-compression).
Doing post-compression is attractive, since it saves Ne 1 compression filtrations, where Ne is the number
of transducer elements.
In this thesis the theoretical difference between pre- and post-compression is investigated. It is shown
that an error is introduced when using post-compression with dynamic focusing. The error magnitude
is studied through simulations, string phantom measurements, and in-vivo measurements from the liver.
Further a modified post-compression scheme which compensates for the error is proposed. The modified
post-compression scheme also offers a substantial reduction in computational workload.
Creation date: 2009-06-08
Update date: 2009-06-09