Track: Basic Science and Instrumentation

Session: Research Abstracts: Basic Science and Instrumentation (8:15 AM – 10:15 AM)

Wednesday, March 29, 2017

9:55 AM – 10:05 AM

In Vivo Vector Flow Imaging for a Portable Ultrasound Scanner

Your role: Author

Your role: Presenter

Objectives

Conventional ultrasound velocity estimation requires beam-to-flow angles below 60 degrees, which are difficult to achieve in the presence of parallel or turbulent flow. Vector flow imaging (VFI) overcomes this limitation by estimating the velocity components in 2- or 3D. Implementing VFI in a tablet-based scanner would be a game changer in the improvement of the current clinical practice.

Methods

A 2D VFI method was recently presented, combining directional transverse oscillation (DTO) with synthetic aperture sequential beamforming. Due to the synthetic aperture, the method provides low standard deviation (3.19% and 0.47% for the lateral and axial components in a vessel at 90 degrees) with a frame rate of 1500 fps. A low bias of –0.67% and –0.43% in the lateral and axial directions is achieved due to the auto-calibration of the beamformers proper of the DTO. The method is computationally inexpensive, based on a phase-shift estimation approach. Furthermore, the wireless transmission of the ultrasound data is enabled by the dual-stage beamformer.

Results

In vivo images of the carotid bifurcation are shown, acquired with a frame rate of 1500 fps. The method enables estimating velocities in a high range, therefore both systole and diastole can be imaged with a single setup. The high frame rate enables the visualization of complex flow patterns. Ten cardiac cycles are measured to evaluate the precision of the velocity estimation.

Conclusions

The in vivo study shows the potential of the proposed method, which enables the implementation of VFI in a tablet-based scanner with high performance and high frame rate.

Learning Objectives:

Vector flow imaging (VFI) methods do not require any manual angle adjustment, and therefore reduce the variability of quantitative flow measurements. In this presentation, a low-complexity, high-frame-rate VFI method is proposed for the implementation in a hand-held ultrasound scanner. In-vivo images from a common carotid artery are displayed to show the potential of the approach. The method is able of estimating velocities in a high range, therefore both the systolic and diastolic phases can be correctly imaged with a single setup. The ultrasound data can be transmitted wirelessly, and the low computational demand enables the possibility of real-time processing in a consumer level tablet.

Role: Author

Tommaso Di Ianni, MSc

PhD student

Technical University of Denmark

Email: todiian@elektro.dtu.dk

Office Telephone: +4545255256

Role: Presenter

Tommaso Di Ianni, MSc

PhD student

Technical University of Denmark

Email: todiian@elektro.dtu.dk

Office Telephone: +4545255256

Role: Author

Jørgen Arendt Jensen, PhD

Professor of Biomedical Signal Processing

Technical University of Denmark

Email: jaj@elektro.dtu.dk

Office Telephone: 004545253898