Fig. 4

The 3D FD (between 2 and 3) of the grayscale cine is computed using the differential box-counting algorithm that takes 3D pixel intensity information into account. In the standard box-counting method applied to binary images as either outlines or filled silhouettes, intensity information is lost as foreground pixels are contrasted from the background pixels to derive the 2D FD (range 1 – 2). For the same original image and considering only the mantissa, it is usually the case that the binary FD is greater than the grayscale equivalent. Furthermore, the FD of the filled binary mask would usually be nearer to 2 when compared to the FD of the equivalent binary outline as the FD of the filled areas massively outweigh the FD of the edges. Abbreviations as in Fig. 2