DICOM software on CUDA
Updated:
Fastvideo DICOM software runs medical image processing on the GPU for PACS and imaging workflows: fast JPEG and JPEG2000 encoding and decoding at up to 16 bits per channel, together with resize, denoising, and rotation. Because the codecs and the whole pipeline run on CUDA, large studies — CT, MR, X-ray, mammography, ultrasound, and angiography — are quick to convert, transfer, and store.
Looking for the underlying codec? See the 12-bit JPEG codec and the JPEG2000 codec. This page is the DICOM software, viewer, and PACS pipeline on CUDA.
There is often the need to use these images in conferences or in presentations; however, the DICOM (Digital Imaging and Communications in Medicine) format that most of these images cannot be used by presentation software such as PowerPoint (Microsoft, Redmond, Wash) and many radiologists lack the ability to manipulate these images on their desktop computers. In addition, DICOM images are quite large and converting these files to a smaller format can improve transfer speed and decrease storage requirements. Some of the picture archiving and communication systems (PACS) available, offer the ability to convert DICOM images into other formats, but for those who lack this ability, there is an assortment of free software available on the Internet that can allow the viewing and manipulating of these images on a personal computer.
Our Image & Video Processing SDK could be utilized in PACS for fast processing of DICOM images. Apart from fast DICOM encoding and decoding we offer other image processing features like resizing, denoising, rotations, etc.
CUDA pipeline for DICOM images
How does GPU DICOM image processing work?
- Image processing for bit depth up to 16 bits per channel
- Fast processing of image series and still images with arbitrary resolution
- DICOM encoding and decoding:
- Lossless JPEG (2-16 bits)
- Baseline JPEG (8/24 bits)
- Extended JPEG (12/36 bits)
- JPEG2000 (1-16 bits, lossless and lossy)
- MPEG4-H.264
- MPEG4-H.265
- Support of the latest NVIDIA GPUs
- Compatibility with Windows-7/8/10, Linux Ubuntu, CentOS, L4T
Modern medical applications are working via browsers, that's why fast DICOM decoding is essential. It could hardly be implemented as a browser plugin, especially for JPEG2000 decoding, because it's very slow on GPU. To solve that problem, we could offer high performance J2K transcoding to JPG on GPU. JPG format is native to all browsers and such a transcoding could be extremly fast.
Apart from J2K to JPG transcoding on GPU we could develop any other custom solutions for DICOM applications.
About the author
Fyodor Serzhenko, PhD, is the founder and CEO of Fastvideo. He earned his PhD at the Moscow Institute of Physics and Technology (MIPT) in 1993. Since 2009 he has led the development of Fastvideo’s GPU-accelerated image codecs and ISP modules, including the Fastvideo SDK. Connect on LinkedIn.
Why you can trust these results
Fastvideo has built GPU-accelerated image processing software since 2009. Our codecs comply with the standards they implement — JPEG (ITU-T T.81 / ISO IEC 10918) and JPEG2000 (ITU-T T.800 / ISO IEC 15444) — and our tools are open source on GitHub. Every performance figure we publish is reproducible: see our benchmark methodology, download the Fastvideo SDK benchmark report (PDF), and measure it on your own GPU.