Technical Note: Virtual phantom analyses for preprocessing evaluation and detection of a robust feature set for MRI‐radiomics of the brain

Abstract

Purpose: The purpose of the paper was to use a virtual phantom to identify a set of radiomic features from T1-weighted and T2-weighted magnetic resonance imaging (MRI) of the brain which is stable to variations in image acquisition parameters and to evaluate the effect of image preprocessing on radiomic features stability. Methods: Stability to different sources of variability (time of repetition and echo, voxel size, random noise and intensity non-uniformity) was evaluated for both T1-weighted and T2-weighted MRI images. A set of 107 radiomic features, accounting for shape and size, first order statistics, and textural features was used. Feature stability was quantified using intraclass correlation coefficient (ICC). For each source of variability, stability was evaluated before and after preprocessing (Z-score normalization, resampling, gaussian filtering and bias field correction). Features that have ICC > 0.75 in all the analysis of variability are selected as stable features. Last, the robust feature sets were tested on images acquired with random simulation parameters to assess their generalizability to unseen conditions. Results: Preprocessing significantly increased the robustness of radiomic features to the different sources of variability. When preprocessing is applied, a set of 67 and 61 features resulted as stable for T1-weighted and T2-wieghted images respectively, over 80% of which were confirmed by the analysis on the images acquired with random simulation parameters. Conclusion: A set of MRI-radiomic features, robust to changes in TR/TE/PS/ST, was identified. This set of features may be used in radiomic analyses based on T1-weighted and T2-weighted MRI images.

Baseline MRI-Radiomics Can Predict Overall Survival in Non-Endemic EBV-Related Nasopharyngeal Carcinoma Patients

“…Only the features that were known to be stable to changes in image acquisition parameters and to geometrical transformation of the ROI were kept. Two experiments to test stability were performed as described in previous studies [ 31 , 44 ]. In the first experiment [ 44 ], radiomic features were extracted from multiple virtual MRI acquisitions of the same phantom to assess the stability to variations in image acquisition parameters such as time of repetition (TR), time of echo (TE) and voxel size.…”

“…Two experiments to test stability were performed as described in previous studies [ 31 , 44 ]. In the first experiment [ 44 ], radiomic features were extracted from multiple virtual MRI acquisitions of the same phantom to assess the stability to variations in image acquisition parameters such as time of repetition (TR), time of echo (TE) and voxel size. In the second experiment [ 31 ], a stability analysis to small translation of the ROI was performed as a surrogate of stability to multiple segmentations.…”

Standardization of brain MR images across machines and protocols: bridging the gap for MRI-based radiomics

“…For example, it has been shown that bias field correction efficiently minimizes MR intensity inhomogeneity within a tissue region 8 – 10 . The variability generated by different voxel sizes can also be reduced by spatial resampling 9 , 11 , 12 . Moreover, brain extraction is mandatory to remove the skull regions that generate the most important variations in intensities and to define the region in which intensities should be considered before any image intensity normalization 13 , 14 .…”

Standardization of imaging methods for machine learning in neuro-oncology

“…These include noise filtering, inhomogeneity correction, and intensity normalization. When preprocessing steps are applied to MRI dataset, the robustness of radiomic features has improved based on phantom analyses 26 as well as data from patients with glioblastoma. 27 …”