PyRadiomics

PyRadiomics is an open-source Python library for the extraction of radiomic features from medical images. It provides a standardized framework to transform imaging data and their associated segmentations into quantitative feature vectors for radiomics analyses, enabling reproducible, high-throughput feature computation.

The library computes a broad set of feature classes, including first-order statistics, shape-based descriptors, and texture features such as GLCM, GLRLM, GLSZM, NGTDM, and GLDM. Features can be calculated for 2D, 3D, or multi-ROI analyses within a study. PyRadiomics supports common imaging modalities including CT, MRI, and PET, and offers preprocessing options such as resampling to isotropic voxel spacing, intensity discretization, normalization, and interpolation. It relies on widely used Python packages (such as NumPy, SciPy, and SimpleITK) and can be used via a Python API or a command-line interface. Input consists of an image and a binary segmentation mask (or label map); the output is a tabular dataset with one row per region of interest and columns for each feature, plus metadata.

Standardization and reproducibility are central goals: PyRadiomics implements feature definitions aligned with the Image Biomarker Standardisation Initiative (IBSI) where available, promoting comparability of results across platforms and studies. Availability and licensing: PyRadiomics is open-source software distributed under the BSD 3-Clause license and is actively maintained by a community of researchers. Documentation, tutorials, and example workflows are hosted on its GitHub page. Use cases include oncology radiomics studies for prognosis, treatment response assessment, and lesion characterization, where standardized feature extraction supports predictive modeling and hypothesis testing. The project began to gain prominence after its initial release in 2017 and has since evolved with ongoing community contributions.