vbrename

vbrename takes a directory full of arbitrarily named DICOM files and sorts them into subdirectories based on various pieces of header information. It also renames the individual files, using timestamps where possible to make sure the lexical order is also the order of acquisition. vbrename does not modify the contents of the files. Here's the help text:

VoxBo vbrename (v1.8.5/Jun 14 2011)
summary:
  sorts DICOM files into separate directories
usage:
  vbrename [flags] <file/dir> [<file/dir> ...]
flags:
  -n       don't remove identifying fields
  -h       show help
  -v       show version
notes:

  vbrename will take a heap of mixed DICOM files and sort them into
  directories based on the series and protocol, and rename the files
  according to instance, date, and time, as in the following:

    0001_localizer/0001_20101019_091004.170008

  By default, vbrename de-identifies each file in the process, using
  the default behavior of dcmsplit, removing:

    all of groups 0010, 0012, 0032, and 0038
    any element with value representation of PN (person name)

  Note that elements embedded inside SQ (sequence) elements are not
  stripped, unless the SQ itself is stripped.

  If you don't want to de-identify your data, use the -n flag -- the
  files will be sorted and renamed, but not modified in any way.  If
  you want more control over how files are stripped, use dcmsplit.

  Note that vbrename will also attempt to make the file and its
  containing directory user-readable and user-writable if possible.