ROdiomiX

ROdiomiX computes standardized radiomic features from medical images to provide reproducible quantitative biomarkers for radiation oncology research.

Key Features:

• Implementation: Implemented in MATLAB.
• Standards compliance: Adheres to the Image-Biomarker-Standardization-Initiative (IBSI) definitions and guidelines.
• Feature categories: Computes 11 radiomic feature categories: Local-Intensity, Intensity-Histogram, Intensity-Based-Statistical, Intensity-Volume-Histogram, Gray-Level-Co-occurrence, Gray-Level-Run-Length, Gray-Level-Size-Zone, Gray-Level-Distance-Zone, Neighborhood-Grey-Tone-Difference, Neighboring-Grey-Level-Dependence, and Morphological features.
• Aggregation methods: Incorporates 10 different feature aggregation methods as specified by IBSI.
• Validation data: Validated using the IBSI 3D digital phantom and an HFH digital phantom.
• Reproducibility metric: Uses intraclass correlation coefficient (ICC) to assess absolute agreement between computed features and benchmark values, reporting ICC values exceeding 0.997 for all ten feature categories across both phantoms.
• Accuracy: Reports percent differences versus benchmark values of 0.14% ± 0.43% for the IBSI digital phantom and 0.11% ± 0.27% for the HFH digital phantom.

Scientific Applications:

• Radiation oncology studies: Provides standardized radiomic feature extraction for research in radiation oncology.
• Reproducibility and comparability: Enables reproducible and comparable radiomic analyses through IBSI compliance and standardized aggregation methods.
• Benchmarking and validation: Supports quantitative benchmarking of feature extraction algorithms using digital phantoms and ICC/percent-difference metrics.
• Quantitative imaging biomarkers: Supplies precise quantitative features for imaging biomarker research and clinical-practice applications in radiation oncology.

Methodology:

Implemented in MATLAB, ROdiomiX follows IBSI definitions to compute the listed radiomic feature categories, applies 10 IBSI aggregation methods, and validates outputs against the IBSI 3D digital phantom and HFH digital phantom using intraclass correlation coefficient (ICC) and percent-difference comparisons.