Automated motion artifacts removal between cardiac long- and short-axis magnetic resonance images

Maria C. Carminati, Francesco Maffessanti, Enrico G. Caiani

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

Abstract

We aimed at developing and testing an automated method for motion artifacts compensation, to reduce potential misalignment between short-axis (SAX) and two-and four-chamber long-axis (2ch4chLAX) cardiac magnetic resonance (CMR) images that could introduce artifacts in advanced 3D volumetric analysis, thus precluding accurate measurements. Each SAX slice of the CMR dataset is shifted by optimizing normalized cross correlation of pixel intensities at slice intersection with 2ch4chLAX. The algorithm accuracy has been tested in a dedicated phantom study and applied to a clinical dataset consisting of end diastolic (ED) and end systolic (ES) CMR SAX and 2ch4chLAX frames obtained in 10 consecutive patients. The algorithm performance evaluated on the phantom dataset provided the residual displacement error after images correction (range values 0-2.5 mm), with registration errors comparable with the pixel resolution. Application to clinical data, comparing by visual inspection the results with and without correction, resulted in a perceived improvement in 52.9% of the analyzed frames, thus proving feasibility and usefulness of the method as a necessary pre-processing step for volumetric analysis of CMR data in clinical setting.

Original languageEnglish
Title of host publicationComputing in Cardiology
Pages689-692
Number of pages4
Volume39
Publication statusPublished - 2012
Event39th Computing in Cardiology Conference, CinC 2012 - Krakow, Poland
Duration: Sep 9 2012Sep 12 2012

Other

Other39th Computing in Cardiology Conference, CinC 2012
CountryPoland
CityKrakow
Period9/9/129/12/12

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Automated motion artifacts removal between cardiac long- and short-axis magnetic resonance images'. Together they form a unique fingerprint.

Cite this