Reproducibility Package for "Robustness of Quantitative MRI-Derived Morphological Features of Ischemic Stroke Lesions: An Open-Data Study"

This record contains the reproducibility package supporting the manuscript “Robustness of Quantitative MRI-Derived Morphological Features of Ischemic Stroke Lesions: An Open-Data Study”.

The study uses the publicly available ISLES 2022 training dataset. No original MRI images or lesion masks are redistributed in this record.

The package includes analysis scripts, case-selection information for all 250 ISLES 2022 training cases, derived per-case morphological measurements, uniform and spatially heterogeneous perturbation results, equivalent-sphere analyses, dominant-component tracking results, native-space sensitivity analyses, synthetic-geometry validation, bootstrap confidence intervals, and the data used to generate the figures reported in the manuscript.

The primary analysis comprises 190 lesion-positive cases represented on a native isotropic 2 × 2 × 2 mm grid. Native-space sensitivity analyses include all 247 lesion-positive cases.

The computational environment used for the final analyses is documented in the accompanying requirements and environment files. The analysis code is provided under the MIT License; derived data and documentation are provided under CC BY 4.0.

The original ISLES 2022 dataset is described in: Hernandez Petzsche et al., Scientific Data 2022, 9, 762. https://doi.org/10.1038/s41597-022-01875-5.