Application of linear predictive image resizing on thyroid images

In nuclear medicine; image processing is an important task. In Scintigraphic visualization in addition to physiological, mathematical and physical effects, small organs also make clearness of the image difficult to understand. Image enlargement and/or resizing without any artifact is essential in order to detect small lesions. In this paper, the method of image resizing using linear predictor is applied. The performance of proposed method is compared with nearest neighborhood, bilinear and cubic interpolation using homogeneity, contrast, signal to noise ratio (SNR) and correlation of histograms.