Table 2 Diagnostic accuracy of group method of data handling models to discriminate tumour differentiation grade in volumes from different segmentation techniques
Tumour differentiation grade | Segmentation techniques | |||
|---|---|---|---|---|
SUVmax 2.5 | Liver pool | 41% SUVmax | ITK-SNAP | |
Model value (sensitivity %/specificity %/overall accuracy %) | ||||
 1 | 78.6/100.0/82.4 | 100.0/100.0/100.0 | 83.3/100.0/86.7 | 100.0/100.0/100.0 |
 2 | 66.7/75.0/70.6 | 90.9/100.0/93.8 | 87.5/85.7/86.7 | 83.3/100.0/93.8 |
 3 | 81.8/66.7/76.5 | 71.4/88.9/81.3 | 90.0/80.0/86.7 | 80.0/100.0/90.6 |
Model value (C-statistic/root mean square error/F-measure) | ||||
 1 | 0.818/0.457/0.778 | 0.989/0.244/0.943 | 0.961/0.288/0.921 | 0.976/0.244/0.946 |
 2 | 0.672/0.606/0.618 | 0.928/0.349/0.875 | 0.940/0.335/0.887 | 0.975/0.233/0.988 |
 3 | 0.672/0.679/0.540 | 0.967/0.321/0.892 | 0.955/0.301/0.901 | 0.944/0.216/0.991 |