Regarding “Segmentation of heterogeneous or small FDG PET positive tissue based on a 3D-locally adaptive random walk algorithm” By DP. Onoma,

To the editor, We read with interest the paper of Onoma, et al. regarding the development and evaluation of a novel PET image segmentation method based on a random walk algorithm and called 3D-LARW [1]. In order to evaluate 3D-LARW, it was applied to three different datasets: a physical phantom containing homogeneous spheres acquired in a single PET/CT scanner, synthetic heterogeneous tumors generated by blurring and adding noise to a manually defined ground-truth, and 14 clinical tumors (for which the surrogate of truth was provided by manual delineations carried out by two different experts).

Source: Computerized Medical Imaging and Graphics - October 8, 2015 Category: Radiology Authors: Mathieu Hatt, Dimitris Visvikis Source Type: research