Ultrasound thermal mapping based on a hybrid method combining cross-correlation and zero-crossing tracking

Chang Wei Huang, Der Hsien Lien, Ben Ting Chen, Jay Shieh, Po Hsiang Tsui, Chuin Shan Chen, Wen Shiang Chen*

*Corresponding author for this work

Research output: Contribution to journalJournal Article peer-review

9 Scopus citations

Abstract

A hybrid method for estimating temperature with spatial mapping using diagnostic ultrasound, based on detection of echo shifts from tissue undergoing thermal treatment, is proposed. Cross-correlation and zero-crossing tracking are two conventional algorithms used for detecting echo shifts, but their practical applications are limited. The proposed hybrid method combines the advantages of both algorithms with improved accuracy in temperature estimation. In vitro experiments were performed on porcine muscle for preliminary validation and temperature calibration. In addition, thermal mapping of rabbit thigh muscle in vivo during high-intensity focused ultrasound heating was conducted. Results from the in vitro experiments indicated that the difference between the estimated temperature change by the proposed hybrid method and the actual temperature change measured by the thermocouple was generally less than 1°C when the increase in temperature due to heating was less than 10°C. For the in vivo study, the area predicted to experience the highest temperature coincided well with the focal point of the high-intensity focused ultrasound transducer. The computational efficiency of the hybrid algorithm was similar to that of the fast cross-correlation algorithm, but with an improved accuracy. The proposed hybrid method could provide an alternative means for non-invasive monitoring of limited temperature changes during hyperthermia therapy.

Original languageEnglish
Pages (from-to)1530-1540
Number of pages11
JournalJournal of the Acoustical Society of America
Volume134
Issue number2
DOIs
StatePublished - 08 2013

Fingerprint

Dive into the research topics of 'Ultrasound thermal mapping based on a hybrid method combining cross-correlation and zero-crossing tracking'. Together they form a unique fingerprint.

Cite this