Skip to main content

Table 4 Fingerprints construction for inter-patient similarity analysis

From: Methodological framework for radiomics applications in Hodgkin’s lymphoma

 Fingerprint_OneFingerprint_All
Dataset HL, typeNon-R/R + R/RNon-R/R + R/R
Patients,n8585
SubsetLesions, siteNodal or extra-nodalNodal + extra-nodal
Lesions,n85543
Features volume-relatedNameTLG
Volume_mL
Volume_vx
Compacity
GLNUGLRLM
RLNUGLRLM
GLNUGLZLM
ZLNUGLZLM
TLG
Volume_mL
Volume_vx
Compacity
CorrelationGLCM
GLNUGLRLM
RLNUGLRLM
Coarseness NGLDM
Busyness NGLDM
LZEGLZLM
LZHGEGLZLM
GLNUGLZLM
ZLNUGLZLM
ZPGLZLM
Number814
Features non-volume-relatedNameSUVmean
SUVQ2
SUVQ3
Entropy_log10 HISTO
Entropy_log2 HISTO
EnergyHISTO
LGREGLRLM
HGREGLRLM
SRHGEGLRLM
LRLGEGLRLM
LRHGEGLRLM
HGZE LZLM
SUVQ3
Entropy_log10 HISTO
Entropy_log2 HISTO
HGREGLRLM
SRHGEGLRLM
LRHGEGLRLM
HGZEGLZLM
Number127
PCA retained transformed features (mapping volume + non-volume data)2 + 27 + 2
  1. HL Hodgkin’s lymphoma, n number, non-R/R non-relapsing/refractory, PCA principal component analysis, R/R relapsing/refractory. For the full spelling of the feature and matrixes names, please refer to the supplementary material