Skip to main content


Fig. 13 | EJNMMI Research

Fig. 13

From: Multi-modal image analysis for semi-automatic segmentation of the total liver and liver arterial perfusion territories for radioembolization

Fig. 13

An example (patient ID BI02 from training set in our database) of a transaxial view of early and late CBCTs, b the 18F-FDG PET/CT images, as well as, c CBCT/CT-based LPT segmentation and tumor segmentation (from 18F-FDG PET). In this slice, the tumor was segmented in different lobes by the two methods. Comparing the CBCT- and CT-based lobe segmentation for this patient showed that 10% of the liver volume was assigned to different lobes (Vdiff), the ratio of “right lobe to entire liver volume ratio” for these two methods (Rratio) was 1.08 and the same parameter for the left lobe (Lratio) was 0.82. Furthermore, the average and maximum distance between lobe separation planes by these two methods (aHD and mHD) were 2.74 and 16.94 mm. Dosimetry simulation showed that a mean absorbed dose to the left and right lobe when using CBCT-based lobe segmentation were 19 and 46 Gy respectively, when we aimed at delivering 40 Gy to each lobe using the CT-based LPT segmentation

Back to article page