Quantitative analysis of the tomographic thallium-201 myocardial bullseye display: critical role of correcting for patient motion

Abstract

Single photon emission computed tomography (SPECT) myocardial 201TI imaging appears to offer major improvements over planar imaging. Quantitative analysis of the 201TI images appears to offer major advantages over subjective analysis in planar imaging, but the three-dimensional data available in SPECT images requires special approaches to analysis and display. Thus the myocardial "bullseye" display was developed to summarize and analyze the three-dimensional images of the left ventricle in two dimensions. The relative 201TI distribution to each region of the left ventricle of an individual patient can be displayed as the number of s.d.s away from normal that the region falls. We found that patient motion during the 22 min required for SPECT imaging appeared to produce artifactual defects. Thus, computer programs were developed to quantitate motion between consecutive frames of a [201TI] SPECT myocardial imaging study, simulate nonreturning vertical motion in normal patients, and correct the acquired data for motion. Motion as small as 0.5-1.0 pixel (3-6 mm) in the vertical (axial) direction caused artifactual defects in the quantitative bullseye display that resulted in a false-positive rate of up to 40% for a +1.0 pixel shift. Patient motion of magnitude greater than the threshold value for artifact-production (0.5 pixel) occurred at a rate of 10%, and should be corrected before tomographic reconstruction.