Skip to main content

Table 2 Diagnostic accuracy of group method of data handling models to discriminate tumour differentiation grade in volumes from different segmentation techniques

From: Multiparametric quantitative and texture 18F-FDG PET/CT analysis for primary malignant tumour grade differentiation

Tumour differentiation gradeSegmentation techniques
SUVmax 2.5Liver pool41% SUVmaxITK-SNAP
Model value (sensitivity %/specificity %/overall accuracy %)
 178.6/100.0/82.4100.0/100.0/100.083.3/100.0/86.7100.0/100.0/100.0
 266.7/75.0/70.690.9/100.0/93.887.5/85.7/86.783.3/100.0/93.8
 381.8/66.7/76.571.4/88.9/81.390.0/80.0/86.780.0/100.0/90.6
Model value (C-statistic/root mean square error/F-measure)
 10.818/0.457/0.7780.989/0.244/0.9430.961/0.288/0.9210.976/0.244/0.946
 20.672/0.606/0.6180.928/0.349/0.8750.940/0.335/0.8870.975/0.233/0.988
 30.672/0.679/0.5400.967/0.321/0.8920.955/0.301/0.9010.944/0.216/0.991
  1. SUV Standardised uptake value