MCVV phantom datasets (MCVV-SPH, MCVV-RPH): synthetic and anatomy-like phantoms with geometric ground truth for volumetric medical-image measurement

The MCVV (Mingwei Cui / VoxelVision) phantom datasets — self-built phantoms providing geometric ground truth for evaluating volumetric medical-image measurement (distance, angle, path, perimeter). This deposit contains: MCVV-SPH — synthetic phantoms with analytically defined ground truth (large-N internal validity); MCVV-SPH-pilot — a pipeline-validation pilot subset; MCVV-RPH — a real anatomy-like airway-tree phantom referenced to independent vernier-caliper distances. Imaging volumes (NIfTI) are provided with documented reference geometry and provenance, suitable for benchmarking measurement tools against known geometry. Associated analysis code and lightweight measurement records are version-controlled in the source-code repository. Released under CC-BY-4.0. Note on MCVV-SPH: the synthetic-phantom archive is split into 20 parts (MCVV-SPH.zip.part00 through .part19) to accommodate upload limits. After downloading all parts, reassemble with the shell command: cat MCVV-SPH.zip.part* > MCVV-SPH.zip (the wildcard sorts part00–19 in order). Verify the reassembled archive with md5: d2f982947b927e2aa196b225cf9f6316. MCVV-SPH-pilot and MCVV-RPH are single archives and need no reassembly.