Deep Learning Combined with Radiologist’s Intervention Achieves Accurate Segmentation of Hepatocellular Carcinoma in Dual-Phase Magnetic Resonance Images

Figure 8

The example of segmentation by deep fusion network- (DFN-) F and DFN-R in which the Dice similarity coefficient (DSC) value was 0.08 and 0.42, respectively. (a) Original image, (b) ground truth GT, (c) segmentation by DFN-F, and (d) segmentation by DFN-R.

(a)

(b)

(c)

(d)