计算机科学
放射性废物
背景(考古学)
数据采集
同种类的
计算机视觉
实验数据
人工智能
鼓
断层摄影术
跟踪(教育)
伽马能谱学
放射性核素
模型
参考数据
数据挖掘
贝叶斯概率
伽马射线
迭代重建
放射源
校准
环境科学
反问题
模拟
材料科学
作者
Victor J. Casas-Molina,Eric Laloy,Bart Rogiers,Tom Dhaene,Ivo Couckuyt
标识
DOI:10.5281/zenodo.24062456
摘要
Dataset Description This dataset contains experimental X-ray computed tomography (CT) and Angular Segmented Gamma Scanning (ASGS) measurements of radioactive waste drum mockups, collected for research into three-dimensional radionuclide activity reconstruction. The measurements were used in the development and evaluation of SARUMAN (Spatial Activity Reconstruction Using Machine-learning Algorithms for Nuclear Waste), a Bayesian machine-learning framework for reconstructing spatial activity distributions from gamma spectrometry and CT-derived attenuation information. Dataset Contents The dataset comprises nine experimental configurations: Six homogeneous mockups (HOM-1 to HOM-6), containing different source arrangements within homogeneous filling materials, including bark and potting soil. Three heterogeneous mockups (HET-1 to HET-3), designed to investigate activity reconstruction under more complex material and source distributions. Each configuration is provided as an individual ZIP archive, containing: CT/: A reconstructed X-ray CT volume in MetaImage format (reconstructionFloat.mhd and reconstructionFloat.raw). ASGS/: Gamma spectrometry acquisition files (.asr, .asw, and .srf) corresponding to the segmented angular measurements. The original spectrometry file formats are preserved to retain the experimental acquisition information. Intended Use These measurements are provided to support reproducibility, methodological comparisons, and further research into radiological waste characterization, gamma-ray transport modelling, uncertainty quantification, and Bayesian inverse problems. The dataset can be used with SARUMAN or adapted for other reconstruction and data-analysis approaches. Reference Publication Users are strongly encouraged to read the associated scientific publication before using or interpreting this dataset: https://www.sciencedirect.com/science/article/pii/S0306454926007383 The publication provides the scientific context for the experimental measurements, the associated reconstruction methodology, and the considerations relevant to their interpretation. Researchers wishing to reproduce the original analyses or adapt the methodology to other applications should consult this publication alongside the SARUMAN source code and documentation. Scope and Limitations The measurements originate from a specific experimental characterization system. Their interpretation requires appropriate knowledge of the detector geometry, acquisition configuration, calibration procedures, and CT attenuation modelling. The associated SARUMAN implementation includes detector-efficiency and CT attenuation models developed for this experimental setup. These models should not be assumed valid for different instruments without appropriate adaptation and validation. The dataset is provided for scientific research purposes and does not constitute a universally validated characterization benchmark. Associated Software SARUMAN — Spatial Activity Reconstruction Using Machine-learning Algorithms for Nuclear Waste https://github.com/vcasasmo/SARUMAN Dataset preparation and software development: Victor J. Casas Molina.
科研通智能强力驱动
Strongly Powered by AbleSci AI