DSB seminar: 4D ultrasound vector flow imaging for intraventricular flow assessment

Morten Smedsrud Wigen will present his PhD work that he has done at the Department of Circulation and Medical Imaging at NTNU.


Abstract: Ultrasound is the most used modality for cardiac assessment due to its availability, low cost and real-time display. It is widely used for both anatomical imaging and for quantification of blood flow and tissue velocities. Intracardiac blood flow measurements are, however, limited by the intrinsic properties of Doppler-based estimation. The only available option today for time- resolved volumetric three-directional blood flow imaging of the heart, is phase-contrast magnetic resonance imaging (PC-MRI), which is a time demanding and expensive modality with limited accessibility. This work investigates vector flow imaging (VFI) which aim to resolve the angle-dependencies inherent in conventional blood flow estimation, applied in three dimensional echocardiography.


Technological advancements in modern ultrasound systems have allowed more processing capabilities, and enabled new and improved imaging strategies. In this work, so-called multiple line acquisition and matrix-array transducers, have been technological enablers, by allowing acquisition in three dimensions and processing of high data rates.


The aim of this thesis is to develop and validate a 4D VFI method applied on the adult left ventricle, utilizing the capabilities of state-of-the-art clinical ultrasound systems. The clinical aim of measuring complex blood flow is to prevent or slow development of congestive heart failure, by early detection of pathological changes in the intraventricular flow circulation.


An adapted speckle tracking method was developed for this purpose, as a hybrid between conventional Doppler-based estimation for radial estimates, in combination with a more computational demanding block-matching approach for the lateral directional estimates. The method was validated using simulations, an in-vitro setup and with PC-MRI. The proposed method proved to work well under good imaging conditions, but is still limited by challenges especially related to filtering, needed for ultrasound blood flow estimation. This is in addition to general ultrasound imaging challenges related to patient and operator variability. Hence, further validation studies are needed to map the clinical utility of the method. Nonetheless, the results contained in this work was promising for ultrasound as a future modality to assess complex 4D intracardiac blood flow.

Published Jan. 9, 2019 10:45 AM - Last modified Jan. 9, 2019 10:45 AM