Awesome

libdicom

libdicom is a C library and a set of command-line tools for reading DICOM WSI files. It is free (MIT licensed), fast, cross-platform, uses little memory, has no dependencies, includes API documentation, and is easy to use from languages like Python.

libdicom returns compressed frame data, not RGB pixel arrays. OpenSlide 4.0 uses libdicom to implement DICOM support and is a better choice if you want to process image files.

libdicom aims to support most popular DICOM WSI variants. If you have a sample file which does not work well, please open an issue and we'll try to add support.

Building from source

cd libdicom-1.1.0
meson setup builddir --buildtype release
meson compile -C builddir
meson install -C builddir

See the installation documentation for build dependencies and installation options.

Sample code

See the documentation for full details.

#include <stdlib.h>
#include <dicom/dicom.h>

int main() {
    const char *file_path = "/path/to/file.dcm";
    DcmError *error = NULL;

    DcmFilehandle *filehandle = dcm_filehandle_create_from_file(&error, file_path);
    if (filehandle == NULL) {
        dcm_error_log(error);
        dcm_error_clear(&error);
        return 1;
    }

    const DcmDataSet *metadata =
        dcm_filehandle_get_metadata_subset(&error, filehandle);
    if (metadata == NULL) {
        dcm_error_log(error);
        dcm_error_clear(&error);
        dcm_filehandle_destroy(filehandle);
        return 1;
    }

    const char *num_frames;
    uint32_t tag = dcm_dict_tag_from_keyword("NumberOfFrames");
    DcmElement *element = dcm_dataset_get(&error, metadata, tag);
    if (element == NULL ||
        !dcm_element_get_value_string(&error, element, 0, &num_frames)) {
        dcm_error_log(error);
        dcm_error_clear(&error);
        dcm_filehandle_destroy(filehandle);
        return 1;
    }

    printf("NumerOfFrames == %s\n", num_frames);

    dcm_filehandle_destroy(filehandle);

    return 0;
}

Or in Python:

import sys
import pylibdicom

file = pylibdicom.Filehandle.create_from_file(sys.argv[1])
metadata = file.get_metadata_subset()
num_frames_tag = pylibdicom.Tag.create_from_keyword("NumberOfFrames")
num_frames = int(metadata.get(num_frames_tag).get_value()[0])
for frame_number in range(1, num_frames + 1):
    frame = file.read_frame(frame_number)
    value = frame.get_value()
    print(f"frame {frame_number} -> {frame} {len(value)} bytes")

# you can also read frames by (x, y) tile position ... this works for
# TILED_FULL and for sparse images
frame = file.read_frame_position(2, 7)
value = frame.get_value()
print(f"frame {2, 7} -> {frame} {len(value)} bytes")

This will print:

$ ./read-frames.py sm_image.dcm
opening libdicom ...
init for libdicom ...
libdicom version: 1.1.0
frame 1 -> <10x10 pixels, 8 bits, 3 bands, RGB> 300 bytes
frame 2 -> <10x10 pixels, 8 bits, 3 bands, RGB> 300 bytes
frame 3 -> <10x10 pixels, 8 bits, 3 bands, RGB> 300 bytes
...

Command-line tools

libdicom comes with two small command-line tools which can be useful for testing.

dcm-dump will print all metadata from a DICOM file. It's fast, and can dump DICOM files of any size while using only a small amount of memory.

For example:

$ dcm-dump sm_image.dcm
===File Meta Information===
(0002,0001) FileMetaInformationVersion | OB | 2 | 1 | 00 01
(0002,0002) MediaStorageSOPClassUID | UI | 30 | 1 | 1.2.840.10008.5.1.4.1.1.77.1.6
(0002,0003) MediaStorageSOPInstanceUID | UI | 64 | 1 | 1.2.826.0.1.3680043.9.7433.3.12857516184849951143044513877282227
(0002,0010) TransferSyntaxUID | UI | 20 | 1 | 1.2.840.10008.1.2.1
(0002,0012) ImplementationClassUID | UI | 28 | 1 | 1.2.826.0.1.3680043.9.7433.1
(0002,0013) ImplementationVersionName | SH | 14 | 1 | wsiget v0.0.1
...

dcm-getframe will read a single frame from a DICOM file.

For example:

dcm-getframe -o tile.raw data/test_files/sm_image.dcm 12

To read frame 12.

Thanks

Development of this library was supported by NCI Imaging Data Commons, and has been funded in whole or in part with Federal funds from the National Cancer Institute, National Institutes of Health, under Task Order No. HHSN26110071 under Contract No. HHSN261201500003l.