Impact of image filtering and assessment of volume-confounding effects on CT radiomic features and derived survival models in non-small cell lung cancer

Abstract

Background: No evidence supports the choice of specific imaging filtering methodologies in radiomics. As the volume of the primary tumor is a well-recognized prognosticator, our purpose is to assess how filtering may impact the feature/volume dependency in computed tomography (CT) images of non-small cell lung cancer (NSCLC), and if such impact translates into differences in the performance of survival modeling. The role of lesion volume in model performances was also considered and discussed.Methods: Four-hundred seventeen CT images NSCLC patients were retrieved from the NSCLC-Radiomics public repository. Pre-processing and features extraction were implemented using Pyradiomics v3.0.1. Features showing high correlation with volume across original and filtered images were excluded.Cox PH with LASSO regularization and CatBoost models were built with and without volume, and their concordance (C-) indices were compared using Wilcoxon signed-ranked test. The Mann Whitney U test was used to assess model performances after stratification into two groups based on low-and high-volume lesions.Results: Radiomic models significantly outperformed models built on only clinical variables and volume.However, the exclusion/inclusion of volume did not generally alter the performances of radiomic models.Overall, performances were not substantially affected by the choice of either imaging filter (overall C-index 0.539-0.590 for Cox PH and 0.589-0.612 for CatBoost). The separation of patients with high-volume lesions resulted in significantly better performances in 2/10 and 7/10 cases for Cox PH and CatBoost models, respectively. Both low-and high-volume models performed significantly better with the inclusion of *, affiliation at the time of the study.

Blood- and Imaging-Derived Biomarkers for Oncological Outcome Modelling in Oropharyngeal Cancer: Exploring the Low-Hanging Fruit

“…Our results also show that clinically relevant features were more often selected from preprocessed images than from the original image. This is coherent with the use of image filtering as a strategy to enhance image properties and to unveil otherwise undetectable information, as previously observed by our group [ 23 ]. Specifically, different permutations of the wavelet filter seem to yield the highest informative content as compared with other modalities (e.g., gradient, lbp-3D, and log-sigma).…”

“…Admittedly, this strategy has been quite unexplored in HNC radiomics. We believe its application to wider datasets may contribute to the full exploitation of radiomic potentials in this clinical setting and help lay the foundation for a more solid methodological basis, as it is being realized in the context of other cancer types, such as the lung [ 23 , 47 , 48 ].…”

“…Radiomic features were extracted from the non-contrast enhanced CTs using Pyradiomics v3.01 in Python 3.7.10 with the use of Numpy 1.19, SimpleITK 2.0, and PyWavelet 1.1. All features and all image types (i.e., image preprocessing filters) were enabled on all CT scans, as previously reported [ 23 ].…”

Awaiting the support of artificial intelligence in lung cancer

“…In a recent study published on Translational Lung Cancer Research , Volpe et al present us the “ Impact of image filtering and assessment of volume - confounding effects on CT radiomic features and derived survival models in non - small cell lung cancer ”. The paper is an explorative modeling study with particular importance suggesting mindfulness in using radiomics to evaluate small (volume) lung nodules ( 6 ).…”

“…The impact of image filtering on the volume dependence and prognostic value of radiomic features has been assessed with the use of the non-small cell lung cancer (NSCLC) open-source radiomic database and tools (pyradiomics and python model implementation). The use of a publicly available database and tools can be considered advantageous as it is a proof-of-concept study ( 6 ).…”