Skip to main content
. 2021 Jul 19;11(7):1296. doi: 10.3390/diagnostics11071296

Table 5.

Results of dimension reduction and predictive performance of the VOIvital-tumour, VOIgross-tumour and combined model for the noradrenergic biochemical profile in the PPGL cohort. For each model, 3 features were selected corresponding to the factors with the highest loadings. AUCs of the radiomic models and the mean AUC of the sham experiment are reported. Features marked with * are not significantly different between delineation methods in the PPGL dataset. KMO: Kaiser–Meier–Olkin (KMO) measure, AUC: area under the receiver operating characteristic curve.

VOIvital-tumour VOIgross-tumour Combined
Features retained after filtering (τ = 0.95) 53/105 57/105 103/210
KMO 0.970 0.969 0.986
Cumulative variance 0.63 0.58 0.53
Selected features
  • -

    First order Minimum

  • -

    Shape Surface Area *

  • -

    GLCM Informational Measure of Correlation 2
  • -

    Shape Surface Area *

  • -

    NGTDM Complexity

  • -

    GLDM Dependence Entropy
  • -

    VOIgross-tumour GLCM Sum entropy

  • -

    VOIvital-tumour Shape maximum 3D diameter *

  • -

    VOIvital-tumour Shape Surface Volume ratio
AUC (95% CI) 0.829 (0.677–0.981) 0.803 (0.640–0.967) 0.791 (0.618–0.963)
Mean AUC sham experiment (100 iterations) 0.643 0.660 0.666