Correlation of cardiac MRI and PET images using lung cavities as landmarks

R. Kim*, T. Aw, S. Bacharach, R. O. Bonow

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

Magnetic resonance images (MRI) are overlapped with positron emission tomography (PET) images to allow a correlation between structure and function of the heart. This is done by using the structural information inherent in PET attenuation images which are themselves used to form the final corrected PET metabolic image. The lungs are chosen as the structural anchor, but first multiple image processing steps are necessary in order to provide consistent edge enhanced lung cavities of both the MRI and PET attenuation images. Then the regions of interest, the lungs, are demarcated via grey-level contour mapping and binary images obtained (between lung and background). The two sets of binary lung images are approximately overlapped using scale invariant first moments, and then precisely matched using iterative translational and rotational cross-correlation functions. The transaxial cut that produces the minimum area of overlap error between the two sets of images gives the best overlap possible, and the 'best fit' original PET metabolic image can be overlayed on top of the MRI structural image. From this process, cardiac MRI and PET images can be aligned to correlate structure and function.

Original languageEnglish (US)
Title of host publicationComputers in Cardiology
PublisherPubl by IEEE
Pages49-52
Number of pages4
ISBN (Print)0818622253
StatePublished - May 1 1991
EventComputers in Cardiology - Proceedings - Chicago, IL, USA
Duration: Sep 23 1990Sep 26 1990

Other

OtherComputers in Cardiology - Proceedings
CityChicago, IL, USA
Period9/23/909/26/90

ASJC Scopus subject areas

  • Software
  • Cardiology and Cardiovascular Medicine

Fingerprint

Dive into the research topics of 'Correlation of cardiac MRI and PET images using lung cavities as landmarks'. Together they form a unique fingerprint.

Cite this