general options

  -h    --help


          print this help text and exit


        --version


          print version information and exit


        --list-decoders


          list transfer syntaxes of decoders and exit


        --arguments


          print expanded command line arguments


  -q    --quiet


          quiet mode, print no warnings and errors


  -v    --verbose


          verbose mode, print processing details


  -d    --debug


          debug mode, print debug information


  -ll   --log-level  [l]evel: string constant


          (fatal, error, warn, info, debug, trace)


          use level l for the logger


  -lc   --log-config  [f]ilename: string


          use config file f for the logger


  +v    --verbose-pc


          show presentation contexts in verbose mode

input options

input file format:


  +f    --read-file


          read file format or data set


  +fo   --read-file-only


          read file format only (default)


  -f    --read-dataset


          read data set without file meta information
input files:


  +rd   --read-from-dicomdir


          read information on input files from DICOMDIR


  +sd   --scan-directories


          scan directories for input files (dcmfile-in)


  +sp   --scan-pattern  [p]attern: string (only with --scan-directories)


          pattern for filename matching (wildcards)


          # possibly not available on all systems


  -r    --no-recurse


          do not recurse within directories (default)


  +r    --recurse


          recurse within specified directories

processing options

transfer syntax conversion:


  -dn   --decompress-never


          never decompress compressed data sets


  +dls  --decompress-lossless


          only decompress lossless compression (default)


  +dly  --decompress-lossy


          decompress both lossy and lossless compression
deflate compression level:


  +cl   --compression-level  [l]evel: integer (default: 6)


          0=uncompressed, 1=fastest, 9=best compression
other processing options:


  -nh   --no-halt


          do not halt on first invalid input file


          or if unsuccessful store encountered


  -nip  --no-illegal-proposal


          do not propose any presentation context that does


          not contain the default transfer syntax (if needed)


  -nuc  --no-uid-checks


          do not check UID values of input files

network options

application entity titles:


  -aet  --aetitle  [a]etitle: string


          set my calling AE title (default: DCMSEND)


  -aec  --call  [a]etitle: string


          set called AE title of peer (default: ANY-SCP)
association handling:


  +ma   --multi-associations


          use multiple associations (one after the other)


          if needed to transfer the instances (default)


  -ma   --single-association


          always use a single association
other network options:


  -to   --timeout  [s]econds: integer (default: unlimited)


          timeout for connection requests


  -ta   --acse-timeout  [s]econds: integer (default: 30)


          timeout for ACSE messages


  -td   --dimse-timeout  [s]econds: integer (default: unlimited)


          timeout for DIMSE messages


  -pdu  --max-pdu  [n]umber of bytes: integer (4096..131072)


          set max receive pdu to n bytes (default: 16384)


        --max-send-pdu  [n]umber of bytes: integer (4096..131072)


          restrict max send pdu to n bytes

output options

general:


  +crf  --create-report-file  [f]ilename: string


          create a detailed report on the transfer


          (if successful) and write it to text file f

Typical Usage

A typical use case of dcmsend is to send arbitrary SOP instances that are stored as DICOM files to a storage SCP. The following command does exactly this:

dcmsend --verbose <peer> <port> *.dcm

If the DICOM files are stored in a hierarchy of directories below the directory 'IMAGES', the following command can be used:

dcmsend -v <peer> <port> --scan-directories --recurse IMAGES

It is also possible to specify multiple directories and to combine the aforementioned approaches (using both file and directory names):

dcmsend -v +sd +r <peer> <port> IMAGES_1 IMAGES_2 test.img *.dcm

If the SOP instances are referenced from a DICOMDIR file, option --read-from-dicomdir (or +rd) can be used to send all referenced DICOM files without loading them already for the association negotiation:

dcmsend -v <peer> <port> --read-from-dicomdir DICOMDIR

And again, all of the above approaches can be combined like this:

dcmsend -v +sd +r +rd <peer> <port> IMAGES_1 IMAGES_2 test.img DICOMDIR *.dcm

The default option --read-file-only makes sure that only DICOM files (i.e. the ones with meta-header and magic word 'DICM' after the preamble) are processed. Usually, if processing a whole bunch of files, it is also a good idea not to halt on the first invalid input file or if an unsuccessful store was encountered. This can be accomplished by using option --no-halt. Please note, however, that 'unsuccessful store' does not mean that the DIMSE status of the C-STORE response indicates an error. It means that the C-STORE request could not be sent to the storage SCP.

If more than 128 presentation contexts are needed, which is the maximum number allowed according to the DICOM standard, a new association is started after the previous one has been completed. In cases where this behavior is unwanted, it can be disabled using option --single-association. In addition, whether only lossless compressed data sets are decompressed (if needed), which is the default, or also lossy compressed data sets can be specified using the --decompress-xxx options.

In order to get both an overview and detailed information on the transfer of the DICOM SOP instances, option --create-report-file can be used to create a corresponding text file. However, this file is only created as a final step if the application did not terminate before (with an error).

Scanning Directories

Adding directories as a parameter to the command line only makes sense if option --scan-directories is also given. If the files in the provided directories should be selected according to a specific name pattern (e.g. using wildcard matching), option --scan-pattern has to be used. Please note that this file pattern only applies to the files within the scanned directories, and, if any other patterns are specified on the command line outside the --scan-pattern option (e.g. in order to select further files), these do not apply to the specified directories.

Thus, the third of the above examples will recurse over directories IMAGES_1 and IMAGES_2 and transmit files that are contained in these two folders and all their subfolders (due to option +r). Additionally, dcmsend will transfer 'test.img' and all files with extension 'dcm' from the current working folder. Note that providing directory names without enabling option +sd does not make sense.

DICOM Conformance

Basically, the dcmsend application supports all Storage SOP Classes as an SCU, including private ones. By default, the application checks the SOP Class UID of the DICOM file in order to make sure that only valid SOP instances are sent. With option --no-uid-checks this check can be disabled.

The dcmsend application also supports all Transfer Syntaxes that are defined in the DICOM standard. Private transfer syntaxes can only be used if the UID check is disabled with option --no-uid-checks. Please note, however, that only a limited number of transfer syntaxes are supported for converting them to the default transfer syntax (Implicit VR Little Endian). With option --list-decoders the transfer syntaxes supported natively or by decoders are listed. The output typically looks like the following:

Transfer syntaxes supported natively:
- Little Endian Implicit
- Little Endian Explicit
- Big Endian Explicit
Transfer syntaxes supported by decoders:
- Deflated Explicit VR Little Endian
- JPEG Baseline
- JPEG Extended, Process 2+4
- JPEG Spectral Selection, Non-hierarchical, Process 6+8
- JPEG Full Progression, Non-hierarchical, Process 10+12
- JPEG Lossless, Non-hierarchical, Process 14
- JPEG Lossless, Non-hierarchical, 1st Order Prediction
- JPEG-LS Lossless
- JPEG-LS Lossy (Near-lossless)
- RLE Lossless

Since dcmsend tries to be as simple as possible for the user, by default presentation contexts might be proposed to the SCP that are strictly speaking 'illegal'. This is because, according to the DICOM standard, the SCU always has to propose the default transfer syntax in at least one presentation context associated with each abstract syntax (i.e. SOP class). This requirement is waived if the SCU only has access to the SOP instance in lossy compressed form or if the decompressed pixel data would be too large to be encoded. With option --no-illegal-proposal the strict DICOM-conformant behavior can be enforced, i.e. no possibly illegal presentation context will be proposed but the corresponding SOP instance will be rejected (if needed). Please note, however, that the size of the decompressed pixel data is not checked.

The default transfer syntax for 'Lossless JPEG Compression', 'Lossy JPEG Compression' and so on are not always proposed as also required by the DICOM standard. The same limitation applies to other compression schemes. See DICOM PS 3.5 section 10 for details.

general

EXITCODE_NO_ERROR                         0
EXITCODE_COMMANDLINE_SYNTAX_ERROR         1

input file errors

EXITCODE_CANNOT_READ_INPUT_FILE          20 (*)
EXITCODE_NO_INPUT_FILES                  21
EXITCODE_INVALID_INPUT_FILE              22
EXITCODE_NO_VALID_INPUT_FILES            23

output file errors

EXITCODE_CANNOT_WRITE_OUTPUT_FILE        40 (*)
EXITCODE_CANNOT_WRITE_REPORT_FILE        43

network errors

EXITCODE_CANNOT_INITIALIZE_NETWORK       60
EXITCODE_CANNOT_NEGOTIATE_ASSOCIATION    61
EXITCODE_CANNOT_SEND_REQUEST             62
EXITCODE_CANNOT_ADD_PRESENTATION_CONTEXT 65

(*) Actually, these codes are currently not used by dcmsend but serve as a placeholder for the corresponding group of exit codes.