King's College London

Research portal

Investigation of microbubble detection methods for super-resolution imaging of microvasculature

Research output: Chapter in Book/Report/Conference proceedingOther chapter contribution

J. Brown, K. Christensen-Jeffries, S. Harput, C. Dunsby, M. X. Tang, R. J. Eckersley

Original languageEnglish
Title of host publication2017 IEEE International Ultrasonics Symposium, IUS 2017
PublisherIEEE Computer Society
ISBN (Electronic)9781538633830
DOIs
Publication statusPublished - 31 Oct 2017
Event2017 IEEE International Ultrasonics Symposium, IUS 2017 - Washington, United States
Duration: 6 Sep 20179 Sep 2017

Conference

Conference2017 IEEE International Ultrasonics Symposium, IUS 2017
CountryUnited States
CityWashington
Period6/09/20179/09/2017

King's Authors

Abstract

Super-resolution techniques that localise isolated bubble signals first require detection algorithms to separate the bubble and tissue responses. This work explores the available bubble detection techniques for super-resolution of tumour microvasculature. Pulse inversion (PI), differential imaging (DI) and singular value decomposition (SVD) filtering were compared in terms of the localisation accuracy, precision and contrast to tissue ratio (CTR). Bubble responses were simulated using the Marmottant model. Non-linear propagation through moving and stationary tissue was modelled using k-Wave. The results showed that PI signal was largely independent of flow direction and speed compared to SVD and DI which were less appropriate for lateral motion. At the lowest speeds, the bubble displacement between frames is not sufficient to generate a strong differential signal. SVD is unsuitable for stationary bubbles. For super-resolution of tumour microvasculature, the results suggest that non-linear techniques are preferential.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454