On the origin of the respiratory artefact species: effect of breathing patterns

Full text: To identify the effect of patterns of respiratory motion on myocardial perfusion artefacts. We used an anthropomorphic phantom with horizontal heart (40 MBq Tc-99m in ventricular wall) but no liver/spleen inserts. Cardiac SPECT was acquired on an OptimaAC camera. The phantom was either IMMOBILE, or had NORMAL (2 second expiratory pause), REVERSE (2 second inspiratory pause) or EVEN (expiratory and inspiratory pauses both 1 second) respiratory motion. Images were reconstructed using filtered back projection (FBP) or attenuation correction (AC). 3 evenly spaced vertical line profiles were taken through the middle vertical long axis image. Horizontal width at fifth maximum (HWFM) values for the outer slopes of the anterior and inferior walls and peak profile counts were obtained for each profile. Ratios of HWFM values for each type of motion to the IMMOBILE HWFM were calculated. Inferior wall is more blurred than the anterior wall with the NORMAL respiratory motion. This pattern is diametrically opposite with REVERSE motion, and absent with EVEN motion. Peak profile counts are always lower when moving than when IMMOBILE. AC results were similar. Pattern of respiration determines which ventricular wall is blurred the most. With normal respiration the inferior wall is more blurred than the anterior wall, producing the inferior wall respiratory artefact. Copyright (2003) The Australian and New Zealand Society of Nuclear Medicine Inc