ViSERA is a desktop software to improve usability, reusability, and reproducibility of radiomics techniques in medical imaging. It is a development platform to create reproducible research workflows by connecting different tools. ViSERA is useful for collaborative research projects and for ensuring consistency across different studies, and is user-friendly for different expertise levels, including radiation oncologists, radiologists, physicists & data scientists.
- ViSERA is a Python-based, entirely-revamped upgrade to our original SERA software (Matlab-based).
- ViSERA is an open-source package that enables standardized and reproducible radiomic feature extraction in compliance with the Image Biomarker Standardization Initiative (IBSI 1.0).
- Image filters have also been standardized against IBSI 2.0 by implementing and validating several filter options.
- ViSERA employs a number of popular image processing algorithms (e.g. for registration, segmentation), to create end-to-end standardized radiomics workflows, applicable to different imaging modalities such as CT, MRI, PET and SPECT.
The radiomics module of ViSERA calculates up to 487 IBSI-standardized features, including: 79 first-order features (morphology, statistical, histogram and intensity-histogram features), 272 higher-order 2D features, and 136 3D features. In addition, it can also calculate 10 moment invariant features, that are not included in IBSI. Different subsets of features can be selected, such as the default of 215 features (first-order + higher-order 3D). The features (entirely consistent with our original SERA) have been studied in multi-center radiomics standardization works by the IBSI (paper), the NIH Quantitative Imaging Network (QIN) (paper) and two other benchmarking studies (paper, paper), with excellent performances and consistency.
Access, Questions & Feedback
You can learn more about and freely access ViSERA here.
Please cite the following reference if you publish results with help from SERA:
M. R. Salmanpour, I. Shiri, M. Hosseinzadeh, H. Zaidi, S. Ashrafinia, M. Oveisi, A. Rahmim
ViSERA: Visualized & Standardized Environment for Radiomics Analysis - A Shareable, Executable, and Reproducible Workflow Generator
Proc. IEEE Medical Imaging Conference, 2023.