Skip to main content

Table 3 Correlation analyses between the dosiomics and DLR features using the Spearman’s rho

From: Radiation pneumonitis prediction after stereotactic body radiation therapy based on 3D dose distribution: dosiomics and/or deep learning-based radiomics features

Settinga

No. of features

Total feature pairs

Correlated pairs and featuresb

Ratio of correlations (%)c

A

107:1316

140,812

(25, 6, 19)

(0.02, 5.61, 1.44)

B

(6645, 54, 736)

(4.72, 50.47, 55.93)

  1. aSetting A: the Spearman’s rho |ρ| ≥ 0.8 between the dosiomics and DLR features and B: the Spearman’s rho 0.5 ≤ |ρ| < 0.8 between the dosiomics and DLR features
  2. bFormat (l, m, n): l is the total number of feature pairs that were correlated, m is the number of dosiomics features correlated with DLR features, and n is the number of DLR features correlated with dosiomics features
  3. cFormat (r, rr, rc): r = number of correlations /total number of feature pairs, rr = number of dosiomics features correlated with DLR features/total number of radiomics features, and rc = number of DLR features correlated with dosiomics features /total number of DLR features used