University of Twente Student Theses
High-speed planar ultrasound blood flow velocimetry using deep learning
Huntink, L.M. (2024) High-speed planar ultrasound blood flow velocimetry using deep learning.
|
PDF
17MB |
| Abstract: | Cardiovascular diseases (CVD) are diseases affecting the blood flow through arteries. These types of diseases are the leading cause of death globally and require early diagnostics to reduce the number of global deaths. Ultrasound is a technique used to image in vivo and can be used to detect CVD. Echo particle image velocimetry (echoPIV) is currently the most common method used to estimate planar velocimetry out of ultrasound B-mode images. However, the technique lacks spatial and temporal resolution. This thesis proposes an alternative method where a residual neural network estimates a planar velocity field from five consecutive ultrasound frames. The neural network is trained on simulated ultrasound data containing curved pulsatile and parabolic blood flow. The trained neural network is experimentally compared to echoPIV by an in vitro measurement of contrast-enhanced water flow through a circular tube. The results show that the neural network improves on echoPIV in its computation time by an evaluation time twenty times faster than echoPIV. Another improvement over echoPIV is the estimation of near-wall velocities. Future research should focus on in vivo measurements |
| Item Type: | Essay (Master) |
| Faculty: | TNW: Science and Technology |
| Subject: | 33 physics |
| Programme: | Biomedical Engineering MSc (66226) |
| Link to this item: | https://purl.utwente.nl/essays/104588 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page