Lossless compression for 3D PET
- Univ. Catholique de Louvain, Louvain-la-Neuve (Belgium). Telecommunication Lab.
- CTI PET Systems, Knoxville, TN (United States)
A new adaptive scheme is proposed for the lossless compression of positron emission tomography (PET) sinogram data. The algorithm uses an adaptive differential pulse code modulator (ADPCM) followed by a universal variable length coder (UVLC). Contrasting with Lempel-Ziv (LZ), which operates on a whole sinogram, UVLC operates very efficiently on short data blocks. This is a major advantage for real-time implementation. The algorithm is adaptive and codes data after some on-line estimations of the statistics inside each block. Its efficiency is tested when coding dynamic and static scans from two PET scanners and reaches asymptotically the entropy limit for long frames. For very short 3D frames, the new algorithm is twice more efficient than LZ. Since an ASIC implementing a similar UVLC scheme is available today, a similar one should be able to sustain PET data lossless compression and decompression at a rate of 27 MBytes/sec. This algorithm is consequently a good candidate for the next generation of lossless compression engine.
- OSTI ID:
- 7124043
- Report Number(s):
- CONF-931051--
- Journal Information:
- IEEE Transactions on Nuclear Science (Institute of Electrical and Electronics Engineers); (United States), Journal Name: IEEE Transactions on Nuclear Science (Institute of Electrical and Electronics Engineers); (United States) Vol. 41:4PT1; ISSN 0018-9499; ISSN IETNAE
- Country of Publication:
- United States
- Language:
- English
Similar Records
The V.42bis standard for data-compressing modems
Graphics processing unit-assisted lossless decompression
Related Subjects
62 RADIOLOGY AND NUCLEAR MEDICINE
ALGORITHMS
COMPUTERIZED TOMOGRAPHY
DIAGNOSTIC TECHNIQUES
EFFICIENCY
EMISSION COMPUTED TOMOGRAPHY
IMAGE PROCESSING
MATHEMATICAL LOGIC
OPTIMIZATION
POSITRON COMPUTED TOMOGRAPHY
PROCESSING
REAL TIME SYSTEMS
TOMOGRAPHY