A novel approach for temporal covariance analysis of gated cardiac studies in scintigraphic images

In this paper, we present a novel approach for temporal covariance analysis of gated studies in a sequence of scintigraphic cardiac images taken throughout the cardiac cycle. Indeed, the covariance function is used as a measure of functional similarity for dynamic analysis of the input image sequence. The originality of our approach resides essentially in the denoising and the enhancement of the studied sequence. This is done while using the principal components analysis in order to deduce relatively to each sequence of scintigrahic images, a new sequence of enhanced and denoised images. Thus, given the general objective, which is the deduction of the evolution direction of the ventricles activity relatively to this of the atria, the resulting covariance image is a functional image representing well the regions with different temporal dynamics. Besides, our method presents the advantage of its easy implementation and real-time parametric map construction. To illustrate the efficiency of the proposed approach, the recorded results for 15 sequences of 16 images, are judged as very encouraging by clinical experts of nuclear medicine belonging respectively to Salah-Azeiez hospital (Tunis, Tunisia) and to Haut-Pierre hospital (Strasbourg, France).