Skip to main content

Table 1 Name and description of PET-AS methods used in this study, with references of published work using similar segmentation approaches

From: Evaluation of prognostic models developed using standardised image features from different PET automated segmentation methods

Algorithm Description Key references
AT 3D adaptive iterative thresholding, using background subtraction Jentzen et al. [43], Drever et al. [44]
RG 3D region-growing with automatic seed finder and stopping criterion Day et al. [45]
KM 3D K-mean iterative clustering with custom stopping criterion Zaidi and El Naqa [8]
FCM 3D fuzzy C-mean iterative clustering with custom stopping criterion Belhassen and Zaidi [46]
GCM 3D Gaussian mixture models based clustering with custom stopping criterion Hatt et al. [37]
WT Watershed transform-based algorithm, using sobel filter Geets et al. [47], Tylski et al. [48]