LIDC-Hybrid-100: A VIDS-Compliant Lung CT Nodule Segmentation Dataset

100-subject lung CT nodule segmentation dataset derived from LIDC-IDRI, structured under the Verified Imaging Dataset Standard (VIDS) v1.0. Note on the name: "Hybrid" refers to the combination of legacy LIDC-IDRI source data (real CT DICOMs, real radiologist annotations from the original four-reader LIDC reads) with modern VIDS-compliant restructuring and provenance documentation. The dataset contains no synthetic images and no synthetic annotations — all imaging and labels originate from the LIDC-IDRI study. Contents: Contents: 100 CT imaging volumes (NIfTI) with VIDS-compliant imaging sidecars; 89 consensus segmentation masks from 4-radiologist annotations; per-annotation provenance (annotator, tool, date, QC status); inter-annotator agreement (mean pairwise Dice 0.7765); quality documentation and ML-ready train/val/test splits (70/15/15); 21/21 VIDS Full profile validation PASS. Source: LIDC-IDRI (Armato et al., Medical Physics 2011) via TCIA. Validate with: pip install vids-validator && vids-validate /path/to/dataset Specification: https://vidsstandard.org