DOI RECORD
HEEH-V1: an open-source DICOM research viewer with reproducible quantitative imaging workflows: technical validation
Abstract
Abstract Background Open-source imaging software can support reproducible research, but implemented capability must be distinguished from clinical validation. We describe HEEH-V1, a research and education DICOM viewer, and evaluate selected quantitative workflows using a public CT phantom series and reference software. Results The fully provisioned local test run passed 335 tests with no skips. On one 172-slice CT phantom series, four matched first-order features (mean, standard deviation, minimum and maximum) were compared slice by slice with PyRadiomics 3.0.1: all 688 feature-by-slice pairs had an absolute difference of zero at a 1 x 10^-6 tolerance. All 270 populated IBSI Configuration D reference rows passed in the configured Z-Rad comparison; a separate PyRadiomics mapping assessed 66 rows, all passing, while 204 remained unassessed. Air-region means on all slices were within the project-set +/-30 HU criterion of -1000 HU (maximum deviation 14.39 HU). Export validation accepted 172/172 tested slices, and 11/11 deterministic manifest-tampering probes were rejected. Conclusions The findings support reproducibility of the specified software and phantom workflows. They do not demonstrate clinical performance, full IBSI compliance, general privacy effectiveness, superiority to established platforms, or equivalence beyond the tested features and data.