Transaxial reconstruction with ramp-filtered backprojection is most common. A Butterworth filter window is best suited for myocardial studies, with order number and cutoff frequency to be determined by the operator according to each system. However, other filters are equally acceptable as far as artifacts are not introduced. Iterative reconstruction methods are now available in modern systems and may represent a better alternative. Oblique-angle reorientation is necessary to obtain short-axis, horizontal and vertical long-axis of the left ventricle. Post-reconstruction magnification is recommended, specially for small hearts. Image display should include comparative slices 1-2 pixel width of stress and rest studies presented in the three planes. Short axis tomograms are displayed from apex to base, horizontal long axis with the apex pointing upwards and septal wall to the left, from anterior to inferior wall, and vertical long axis with the apex to the right and the anterior wall above, from septal to lateral wall.
Semi-quantitative evaluation of myocardial perfusion can be achieved through bull's eye display or polar maps, instrumented in most systems. This is a two-dimensional representation of a three-dimensional distribution of the radiotracer in the myocardium which allows visualization of perfusion defects in a compressed format, including their extension and degree of reversibility. Previous processing steps for the generation of polar maps are to be followed carefully as described by the manufacturer in order to avoid introduction of artifacts, and the results should always be compared with direct observation of conventional tomograms.
For gated SPECT, special software is commercially available from at least four different developers, although two of the program packages have gained more widely acceptance: QGSPECT (Cedars Sinai) and Emory Toolbox (Emory University). Processing usually requires minimal operator input and the results are automatically calculated and displayed in a fixed format containing qualitative and quantitative information on ventricular function including left ventricular ejection fraction and volumes. Observation of dynamic data in cine mode is essential to evaluate regional wall motion and thickening in the three orthogonal planes.
If available, attenuation correction should be applied as far as the method employed is reliable and has been validated. Attenuation correction offers the potential for improved diagnostic accuracy but require a modified approach to image interpretation because resultant images may have different "normal" patterns as compared to conventional, non-corrected images. The Society of Nuclear Medicine and the American Society of Nuclear Cardiology have published a joint position statement recommending that attenuation correction should be regarded as a method for which the weight of evidence and opinion is in favour of its usefulness.
