Reliability of tumor segmentation in glioblastoma: Impact on the robustness of MRI‐radiomic features

Abstract

Purpose The use of radiomic features as biomarkers of treatment response and outcome or as correlates to genomic variations requires that the computed features are robust and reproducible. Segmentation, a crucial step in radiomic analysis, is a major source of variability in the computed radiomic features. Therefore, we studied the impact of tumor segmentation variability on the robustness of MRI radiomic features. Method Fluid‐attenuated inversion recovery (FLAIR) and contrast‐enhanced T1‐weighted (T1WICE) MRI of 90 patients diagnosed with glioblastoma were segmented using a semiautomatic algorithm and an interactive segmentation with two different raters. We analyzed the robustness of 108 radiomic features from five categories (intensity histogram, gray‐level co‐occurrence matrix, gray‐level size‐zone matrix (GLSZM), edge maps, and shape) using intra‐class correlation coefficient (ICC) and Bland and Altman analysis. Results Our results show that both segmentation methods are reliable with ICC ≥ 0.96 and standard deviation (SD) of mean differences between the two raters (SDdiffs) ≤ 30%. Features computed from the histogram and co‐occurrence matrices were found to be the most robust (ICC ≥ 0.8 and SDdiffs ≤ 30% for most features in these groups). Features from GLSZM were shown to have mixed robustness. Edge, shape, and GLSZM features were the most impacted by the choice of segmentation method with the interactive method resulting in more robust features than the semiautomatic method. Finally, features computed from T1WICE and FLAIR images were found to have similar robustness when computed with the interactive segmentation method. Conclusion Semiautomatic and interactive segmentation methods using two raters are both reliable. The interactive method produced more robust features than the semiautomatic method. We also found that the robustness of radiomic features varied by categories. Therefore, this study could help motivate segmentation methods and feature selection in MRI radiomic studies.

Radiomics feature reproducibility under inter-rater variability in segmentations of CT images

“…Parmer et al 13 found that features extracted from automatic segmentations had a better reproducibility that those extracted from manual segmentations. Tixier et al 14 have investigated segmentation variability between two raters and manual and semi-automatic segmentation methods for MR images of glioblastoma. They found that variation between two consecutive scans was higher than variation between segmentations for most features.…”

MRI-based radiomics in breast cancer: feature robustness with respect to inter-observer segmentation variability

“…Although MRI feature robustness has already been investigated for different tumor sites ( e.g., cervical cancer 19 and glioblastoma 23 ), the effect of inter-observer variability segmentation is most likely tumor-site specific 38 . The feature groups enclosing the most robust features in previous investigations (shape 19 and, Intensity-histogram and GLCM 23 ) are different from what we found to be the feature group enclosing the most robust features (local intensities and GLRLM). Most likely this could be explained that different tumor sites influence inter-observer variability.…”

Correction for Magnetic Field Inhomogeneities and Normalization of Voxel Values Are Needed to Better Reveal the Potential of MR Radiomic Features in Lung Cancer

“…As all data were acquired in the same institution and using the same scanner, our findings would also need a multi-center validation. Another limitation is due to the fact that only one operator segmented the tumors and the robustness of the findings with respect to the tumor delineation should be further investigated (52). To investigate the impact of the tumor delineation on our results, all the segmented tumors were automatically eroded by an element of size 1.5 mm and results were similar, confirming the greater sensitivity of 2D analysis on normalized N4ITK corrected data to identify discriminant features (see Supplemental Table 5).…”