Data reduction techniques for the analysis and interpretation of dynamic FDG-PET oncological studies

Dynamic positron emission tomography studies produce a large amount of image data, from which clinically useful parametric information can be extracted with the use of tracer kinetic methods. In order to facilitate the initial interpretation and visual analysis of these large image sequences, data reduction methods can be applied which at the same time maintain important information and allow basic feature characterization. We show here that the application of principal component analysis can provide high-contrast parametric image sets of lesser dimension than the original ones separating structures with different kinetic characteristics. This method has been shown to be an alternative quantification method, independent of any kinetic model and particularly useful when the retrieval of the arterial input function is complicated. Furthermore, novel similarity mapping techniques are proposed, which can summarize in a single image the temporary properties of the whole image sequence according to a reference region. Based on the newly introduced cubed sum coefficient similarity measure, we show that structures with similar time activity curves similar to the tumor's ones can be identified, thus facilitating the detection of lesions not easily discriminated with the conventional method using standardized uptake values.