Quantitative cardiac PET imaging: reproducibility in an integrated analysis environment

The strength of PET imaging in nuclear cardiology relies significantly on the usage of physiological contrast agents such as water or glucose analogues in very small doses in order to assess function, perfusion and metabolism. However, one is continually challenged with maintaining a balance between acquisition capabilities, physiological knowledge and applications in patient care. The more recent use of dynamic studies with high temporal resolution or as function of the heart phases has extended the spectrum of functional, non-invasive cardiac imaging. However, this is paralleled with a drastic increase in data and lack of sufficient means of data analysis. In order to cope with these demands of extracting physiological information from vast amounts of physical data, the analysis environment 'MunichHeart' was developed. Three quantitative analysis modules from this tool box are investigated with respect to intra- and interobserver variability: the assessment of myocardial viability, the delineation of absolute myocardial blood flow and the measurement of left ventricular ejection fraction. The results demonstrate that reproducibility, stability and flexibility can be achieved both in a research oriented environment as well as for routine applications. This is the necessary prerequisite to facilitate the transfer of new methods into clinical reality.