HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Martínez-Möller, A. Articles by Nekolla, S. PubMed Articles by Martínez-Möller, A. Articles by Nekolla, S.

Automatic cardiac PET-SPECT and SPECT-SPECT registration

¹ Nuklearmedizinische Klinik der TU München, München, Germany; ² CAMP-AR, TU München, Munich, Germany

121

Objectives: Combination of examinations acquired in rest and stress or the assessment of perfusion and metabolism is a unique capability of nuclear imaging. However, most protocols require that the patient leaves the scanner between examinations or even use different modalities. The alignment of these studies is mostly done manually which introduces a potential bias. Thus we aimed at the development and initial validation of an automated approach suitable for both SPECT and PET.

Methods: Using the ITK toolkit, an automatic rigid image registration using Mattes’ mutual information was implemented and applied in 2 consecutive sets of patients: 12 patients undergoing a rest/stress Tc99m-SPECT and 12 patients undergoing a rest Tc99m-SPECT and an FDG PET for viability assessment. Manual reference registration was performed by an experienced observer. The deviation of the automatic alignment was then compared to the reference. In addition, a second observer blinded to the previous results assessed the visual quality of the algorithm.

Results: The average translation between the reference and the automatic approach in x,y,z was 2.3mm, 3.3mm and 4.9mm for PET-SPECT and 3.1mm, 1.1mm, 1.8mm for SPECT-SPECT. The difference in the rotations was minimal (<1°). In the quality read, 2/12 PET-SPECT and 1/12 SPECT-SPECT registrations were found to need a manual correction (25mm, 19mm, 13mm translations).The average time needed for the automatic registration was 30s for PET-SPECT, 3s for SPECT-SPECT on standard computers. Manual registration required 2-3 min per case.

Conclusions: Automatic registration of cardiac studies is successfully achieved for intra- and intermodal PET and SPECT imaging. This offers the potential of an additional improvement of the clinical workflow with respect to analysis time and reproducibility.

This Article Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Google Scholar Articles by Martínez-Möller, A. Articles by Nekolla, S. PubMed Articles by Martínez-Möller, A. Articles by Nekolla, S.