DICOM¶

VoxelKit handles both single .dcm files and series directories (a folder of per-slice .dcm files forming a 3D volume) through the same set of functions. You point the function at a file or a directory and it figures out which case it is.

Patient identifiers are stripped from inspect_dicom output by default. You have to opt in with include_phi=True to see them, which keeps accidental PHI exposure out of logs and debug prints.

from voxelkit import inspect_dicom, preview_dicom, report_dicom
from voxelkit.dicom import anonymise_directory, dicom_to_nifti

Functions¶

Function	What it does
`inspect_dicom`	Metadata dict from a `.dcm` file or series directory
`preview_dicom`	PNG slice from a `.dcm` file or series
`report_dicom`	QA stats for a `.dcm` file or series
`anonymise_directory`	Scrub PHI tags from a whole directory tree
`dicom_to_nifti`	Convert a `.dcm` file or series to NIfTI

Series detection¶

When you pass a directory to any of the DICOM functions, VoxelKit looks for .dcm files at the top level and assembles them as a single series:

Slices are sorted by ImagePositionPatient[2] when present (patient-relative Z).
Falls back to InstanceNumber, then filename order.
All slices must share the same in-plane shape and dtype. A heterogeneous directory raises ValidationError so you notice if two acquisitions have been mixed together.

For nested directory trees (one folder per study, one per series), call the function once per series folder. There is no cross-folder auto-grouping in this release.