Graph feature selection for enhancing radiomic stability and reproducibility across multiple institutions in head and neck cancer
Authors:
Open Access
- 31 July 2025
- journal article
- research article
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-16
Abstract
Radiomic biomarkers offer promise for precision oncology. However, their clinical utility is limited by variability from differing imaging protocols and the high dimensionality of radiomics data. Feature selection is key for better interpretability, accuracy, and efficiency, yet traditional methods lack stability and reproducibility. We investigate a Graph-Based Feature Selection (Graph-FS) approach that models feature interdependencies to identify stable radiomic signatures for head and neck squamous cell carcinoma (HNSCC) across institutions. We retrospectively analyzed 1,648 radiomic features extracted from the gross tumor volumes of 752 HNSCC patients from three institutions. After standard preprocessing and applying 36 radiomics parameter configurations to simulate variability, we compared Graph-FS with established methods: Boruta, Lasso, Recursive Feature Elimination (RFE), and Minimum Redundancy Maximum Relevance (mRMR). We evaluated feature selection stability and reproducibility using Pearson correlation, the Jaccard Index (JI), and the Dice-Sorensen Index (DSI) and assessed ranking consistency with Kendall’s Coefficient of Concordance (W). Graph-FS achieved higher stability (JI = 0.46, DSI = 0.62, OP = 45.8%) versus baseline methods with JI of 0.005 (Boruta), 0.010 (Lasso), 0.006 (RFE) and 0.014 (mRMR). These results demonstrate that Graph-FS enhances feature stability, reproducibility, and predictive performance. This method could facilitate integration into AI-driven radiomics workflows for reliable, multi-center biomarker discovery.Keywords
Funding Information
- National Institutes of Health (R01EB032680, R01CA279013, R01DE033426, and U54CA273956, R01EB032680, R01CA279013, R01DE033426, and U54CA273956, R01EB032680, R01CA279013, R01DE033426, and U54CA273956, R01EB032680, R01CA279013, R01DE033426, and U54CA273956, R01EB032680, R01CA279013, R01DE033426, and U54CA273956, R01EB032680, R01CA279013, R01DE033426, and U54CA273956)
This publication has 46 references indexed in Scilit:
- Graph neural networks in multi-stained pathological imaging: extended comparative analysis of Radiomic featuresInternational Journal of Computer Assisted Radiology and Surgery, 2024
- Multicentre validation of CT grey-level co-occurrence matrix features for overall survival in primary oesophageal adenocarcinomaEuropean Radiology, 2024
- A new definition for feature selection stability analysisAnnals of Mathematics and Artificial Intelligence, 2024
- Multi-Level fusion graph neural network: Application to PET and CT imaging for risk stratification of head and neck cancerBiomedical Signal Processing and Control, 2024
- Achieving imaging and computational reproducibility on multiparametric MRI radiomics features in brain tumor diagnosis: phantom and clinical validationEuropean Radiology, 2023
- Adaptive unsupervised feature selection with robust graph regularizationInternational Journal of Machine Learning and Cybernetics, 2023
- Minimizing acquisition-related radiomics variability by image resampling and batch effect correction to allow for large-scale data analysisEuropean Radiology, 2020
- Impact of image preprocessing methods on reproducibility of radiomic features in multimodal magnetic resonance imaging in glioblastomaJournal of Applied Clinical Medical Physics, 2019
- Stable feature selection for biomarker discoveryComputational Biology and Chemistry, 2010
- Connected Components in Random Graphs with Given Expected Degree SequencesAnnals of Combinatorics, 2002