We are pleased to announce the release of a new scientific publication by ACMIT and its scientific partners in the journal Scientific Data.
Journal: Scientific Data
DOI: 10.1038/s41597-026-06781-8
Authors: Sepideh Hatamikia, Elisabeth Steiner, Eashrat Jahan Muniya, Soraya Elmirad, Arezoo Borji, Gernot Kronreif, Wolfgang Birkfellner & Martin Buschmann
Radiomics offers valuable quantitative information from medical images but still struggles with poor reproducibility due to differences in imaging devices, acquisition settings, and reconstruction methods. This study addresses that challenge by presenting the first open‑access CBCT phantom dataset designed to test radiomics feature stability in C‑arm linear accelerator imaging. Using a Catphan phantom, the authors acquired 120 CBCT volumes across various devices and imaging parameters, each with ROI segmentations and extracted radiomics features. The dataset aims to improve standardization, support reproducibility testing, and enable the development of more robust radiomics models for clinical use.
This publication marks an important step toward more reproducible and reliable CBCT-based radiomics research.
For more information, please visit:
