Table 1.

CT radiomics in head and neck cancer

Author and year of publication	Sample size and disease stage	Time of imaging and study type	Therapy	Number of parameters derived	Method of feature grouping/selectionand number of parametrs derived	Outcome parameter	Result
M.Bogowicz16 May 2017	N = 93 TC N = 56 VC Stage III and IV HNSCC (Orp,Hyp,Lar,OC)	Pre-treatment retrospective for TC prospective for VC	Definitive IMRT 70 Gy with cisplatin or cetuximab.	Shape Intensity Texture Wavelet transform 317 features	Grouping: PCA Selection: UVCRA for prognosis (9) For comparison with clinical and combined radiomics-clinical model: MVCRA(3) Split ROC curves at 18 mths	Predict LC: using CI Compare radiomics versus clinical model and a combined clinicoradiomic model for LC	Radiomics signature significantly associated with LC Combined Radiomics + Clinical model performed better than Radiomics model alone in TC, but not VC
H. Zhang et alDec17 2013	N = 72 Stage III and IV HNSCC (Orp,Hyp,Lar,OC)	Pretreatment Retrospective Median follow up time: 1.9 years Median OS of entire cohort: 2.6 years	All induction TPF chemotherapy.	First-order texture and histogram analysis, Multiple spatial filters applied from fine to coarse. Number of features = NM	MVCRA for primary mass parameters with OS MVA for model with primary size and N stage. No internal validation/bootstrapping	Predict OS: HR, CI and p-value	MVA: Entropy and skewness with multiple filters associated with OS. MVA of clinical and imaging variables: Entropy and skewness with 1.0 spatial filter associated with OS.
D.Ou et al18 June 2017	N = 120 Stage III-IVb HNSCC	Pretreatment Retrospective Median follow-up time: 49.3  months 5 year actuarial OS: 61.2%	3D-CRT or IMRT with Cisplatin or Cetuximab Matched 2:1 into Concurrent Chemoradiotherapy vs BioRadiotherapy	Shape Intensity Texture Filter based wavelet 544	Grouping: PCA Feature selection: UVCPHA(24) Data dichotomized: low and high radiomics score Internal validation: dichotomizing data and two sided p values	Prognosis and Prediction : OS, PFS AUC at 5 years MVA with HR Test radiomics with combined model using p16	Radiomics model alone, and in combination with p16 predicted OS and PFS. Patients with high signature score significantly benefited more from CRT ( vs BRT) in terms of OS and PFS, while no benefit difference between CRT and BRT in patients with low signature score.

HR, Hazard Ratios; Hyp, Hypopharynx; IMRT, Intensity Modulated Radiotherapy; LC, Local Control; Lar, Larynx; MVA, Multivariate analysis; MVCRA, Multivariate Cox Regression Analysis; N, Number of patients; NM, Not mentioned; OC, Oral cavity; Orp, Oropharnx; PCA, Principal Component Analysis; ROC, Reciever Operated Curves; TC, Training Cohort; TPF, cisplatin, 5-fluorouracil, and docetaxel; UVCRA, Univariate Cox Regression Analysis; VC, Validation Cohort.