# Assessing robustness of radiomic features by image perturbation

arXiv: https://arxiv.org/abs/1806.06719 · HTML: https://mlsanity.com/paper/1806.06719/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Jun 2018.

## Built on by (most cited first, 9 of 9)

- [Random Boxes Are Open-world Object Detectors](https://mlsanity.com/paper/2307.08249/) (2023)
- [RadiomicsFill-Mammo: Synthetic Mammogram Mass Manipulation with Radiomics Features](https://mlsanity.com/paper/2407.05683/) (2024)
- [A Novel Collaborative Self-Supervised Learning Method for Radiomic Data](https://mlsanity.com/paper/2302.09807/) (2023)
- [Tensor Radiomics: Paradigm for Systematic Incorporation of Multi-Flavoured Radiomics Features](https://mlsanity.com/paper/2203.06314/) (2022)
- [Graph data modelling for outcome prediction in oropharyngeal cancer patients](https://mlsanity.com/paper/2310.02931/) (2023)
- [Unraveling Radiomics Complexity: Strategies for Optimal Simplicity in Predictive Modeling](https://mlsanity.com/paper/2407.04888/) (2024)
- [Uncertainty quantification for improving radiomic-based models in radiation pneumonitis prediction](https://mlsanity.com/paper/2412.19511/) (2024)
- [Segmentation variability and radiomics stability for predicting Triple-Negative Breast Cancer subtype using Magnetic Resonance Imaging](https://mlsanity.com/paper/2504.01692/) (2025)
- [Machine Learning based Analysis for Radiomics Features Robustness in Real-World Deployment Scenarios](https://mlsanity.com/paper/2510.25026/) (2025)

## Abstract (arXiv)

Image features need to be robust against differences in positioning, acquisition and segmentation to ensure reproducibility. Radiomic models that only include robust features can be used to analyse new images, whereas models with non-robust features may fail to predict the outcome of interest accurately. Test-retest imaging is recommended to assess robustness, but may not be available for the phenotype of interest. We therefore investigated 18 methods to determine feature robustness based on image perturbations. Test-retest and perturbation robustness were compared for 4032 features that were computed from the gross tumour volume in two cohorts with computed tomography imaging: I) 31 non-small-cell lung cancer (NSCLC) patients; II): 19 head-and-neck squamous cell carcinoma (HNSCC) patients. Robustness was measured using the intraclass correlation coefficient (1,1) (ICC). Features with ICC≥0.90 were considered robust. The NSCLC cohort contained more robust features for test-retest imaging than the HNSCC cohort (73.5% vs. 34.0%). A perturbation chain consisting of noise addition, affine translation, volume growth/shrinkage and supervoxel-based contour randomisation identified the fewest false positive robust features (NSCLC: 3.3%; HNSCC: 10.0%). Thus, this perturbation chain may be used to assess feature robustness.