Related Experiment Videos
Lossless compression of volumetric medical images with improved three-dimensional SPIHT algorithm
Sungdae Cho1, Dongyoun Kim, William A Pearlman
1Centre for Image Processing Research, Department of Electrical, Computer and Systems Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA.
Journal of Digital Imaging
|July 17, 2004
Summary
This study introduces an improved 3-D SPIHT algorithm for lossless compression of volumetric medical images, achieving significant improvements over existing 2-D and 3-D methods.
Area of Science:
- Medical Imaging
- Computer Science
- Signal Processing
Background:
- Volumetric medical image compression is crucial for storage and transmission.
- Existing 2-D and 3-D compression techniques have limitations in efficiency and flexibility.
Purpose of the Study:
- To present an improved three-dimensional (3-D) set partitioning in hierarchical tree (SPIHT) algorithm for lossless compression of volumetric medical images.
- To enhance compression ratios compared to existing 2-D and 3-D methods.
- To introduce flexibility in dimensional decomposition for optimized coding.
Main Methods:
- Utilized an improved 3-D SPIHT algorithm employing asymmetric trees.
- Linked wavelet coefficients generated by 3-D reversible integer wavelet transforms.
- Applied varying decomposition levels between transaxial and axial dimensions.
Main Results:
- Achieved an average improvement of approximately 42% in lossless compression over 2-D techniques.
- Demonstrated superior performance compared to previous 3-D compression techniques.
- Enabled flexible decomposition, beneficial for limited slice coding units.
Conclusions:
- The improved 3-D SPIHT algorithm offers superior lossless compression for volumetric medical images.
- The algorithm provides significant gains over 2-D methods and existing 3-D approaches.
- The adaptable decomposition feature enhances practical applicability in medical imaging scenarios.