Premium
Comparison of robust to standardized CT radiomics models to predict overall survival for non‐small cell lung cancer patients
Author(s) -
Vuong Diem,
Bogowicz Marta,
Denzler Sarah,
Oliveira Carol,
Foerster Robert,
Amstutz Florian,
Gabryś Hubert S.,
Unkelbach Jan,
Hillinger Sven,
Thierstein Sandra,
Xyrafas Alexandros,
Peters Solange,
Pless Miklos,
Guckenberger Matthias,
TanadiniLang Stephanie
Publication year - 2020
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1002/mp.14224
Subject(s) - radiomics , lung cancer , medicine , computed tomography , cancer , medical imaging , radiology , medical physics , nuclear medicine , oncology
Background Radiomics is a promising tool for the identification of new prognostic biomarkers. Radiomic features can be affected by different scanning protocols, often present in retrospective and prospective clinical data. We compared a computed tomography (CT) radiomics model based on a large but highly heterogeneous multicentric image dataset with robust feature pre‐selection to a model based on a smaller but standardized image dataset without pre‐selection. Materials and methods Primary tumor radiomics was extracted from pre‐treatment CTs of IIIA/N2/IIIB NSCLC patients from a prospective Swiss multicentric randomized trial (n patient = 124, n institution = 14, SAKK 16/00) and a validation dataset (n patient = 31, n institution = 1). Four robustness studies investigating inter‐observer delineation variation, motion, convolution kernel, and contrast were conducted to identify robust features using an intraclass correlation coefficient threshold >0.9. Two 12‐months overall survival (OS) logistic regression models were trained: (a) on the entire multicentric heterogeneous dataset but with robust feature pre‐selection (MCR) and (b) on a smaller standardized subset using all features (STD). Both models were validated on the validation dataset acquired with similar reconstruction parameters as the STD dataset. The model performances were compared using the DeLong test. Results In total, 113 stable features were identified (n shape = 8, n intensity = 0, n texture = 7, n wavelet = 98). The convolution kernel had the strongest influence on the feature robustness (<20% stable features). The final models of MCR and STD consisted of one and two features respectively. Both features of the STD model were identified as non‐robust. MCR did not show performance significantly different from STD on the validation cohort (AUC [95%CI] = 0.72 [0.48–0.95] and 0.79 [0.63–0.95], p = 0.59). Conclusion Prognostic OS CT radiomics model for NSCLC based on a heterogeneous multicentric imaging dataset with robust feature pre‐selection performed equally well as a model on a standardized dataset.