Table 2 Radiomic features in the breast cancer risk prediction signature

From: Multicenter radio-multiomic analysis for predicting breast cancer outcome and unravelling imaging-biological connection

 

ROI

Phase

Wavelet

Category

Feature

f1

T

Enh21

HHL

GLCM

Cluster Shade

f2

T

Enh20

HHL

First order

Mean

f3

P

Precontrast

LHL

First order

Energy

f4

P

Precontrast

LHH

First order

Energy

f5

P

Precontrast

HLL

First order

Energy

f6

P

Precontrast

HHH

GLCM

Imc2

f7

P

Skewness

LHH

GLSZM

LGLZE

f8

P

Skewness

LLL

GLDM

GLV

f9

P

Mean

HLH

GLSZM

SALGLE

f10

P

Variance

LHH

GLRLM

SRLGLE

f11

P

Enh20

LHH

GLDM

DE

f12

W

Precontrast

HLH

GLSZM

SALGLE

f13

W

Skewness

LHH

NGTDM

Complexity

  1. T tumor, P peritumor, W tumor and peritumor.
  2. Enh21 = (S2-S1)/S1, Enh20 = (S2-S0)/S0, S0, S1 and S2 represent the precontrast image, early postcontrast image and late postcontrast image series, respectively.
  3. H high-pass filters, L low-pass filters.
  4. LGLZE Low Gray Level Zone Emphasis, GLV Gray Level Variance, SALGLE Small Area Low Gray Level Emphasis, SRLGLE Short Run Low Gray Level Emphasis, DE Dependence Entropy.