DCMfx

Appearance

DICOM Conformance ​

DCMfx reads all valid DICOM Part 10 (P10) data, as well as many DICOM files that that don't strictly conform to the DICOM P10 standard.

When writing DICOM P10 data, strict conformance of the data being written is not enforced, because any DICOM P10 data that was able to be read should also be able to be written, even if parts of it were non-conformant in some way. Still, many variants of invalid DICOM P10 data or files will be corrected by going through a read/write cycle in DCMfx.

UTF-8 Conversion ​

DCMfx converts all strings to UTF-8 when reading DICOM P10 data. This is because native DICOM string data is complex to work with, and UTF-8 is the preferred string encoding of modern systems.

All Specific Character Sets defined by the DICOM standard are supported, including the use of Code Extensions via ISO 2022 escape sequences. If an invalid byte is encountered during UTF-8 conversion then it is converted to the � (U+FFFD) character.

Due to the UTF-8 conversion, DICOM P10 data written by this library always uses the ISO_IR 192 (UTF-8) Specific Character Set.

Sequences and Items of Undefined Length ​

DCMfx converts sequences and items that have defined lengths to use undefined lengths with explicit delimiters. This consumes slightly more space, which may be noticeable for data sets with a very large number of sequences or items, but is necessary in order to be able to stream DICOM P10 data in a memory-efficient way.

DICOM JSON ​

DCMfx supports conversion to and from the DICOM JSON Model. The following details are relevant to users of this feature:

  1. The DICOM JSON specification is optionally extended to allow encapsulated pixel data to be stored. This is done by encoding the binary data present in the '(7FE0,0010) PixelData' data element in Base64. This matches the behavior of other libraries such as pydicom.

  2. BulkDataURI specifiers for storing and retrieving data from external sources are not supported. Binary data must be encoded inline using Base64.

  3. Because floating point Infinity, -Infinity, and NaN are not supported by JSON's number type, DCMfx stores them as quoted strings: "Infinity", "-Infinity", and "NaN".

  4. 64-bit integer values outside the range representable accurately by JavaScript's number type are stored as quoted strings to avoid any loss of precision.