Wavelet coding of volumetric medical datasets

Area of Science:

• Medical Imaging
• Data Compression
• Signal Processing

Background:

• Existing three-dimensional (3-D) discrete cosine transform (DCT) methods for volumetric data coding lack lossless compression and scalability.
• These limitations hinder their application in medical imaging, where high fidelity and flexible data handling are crucial.

Purpose of the Study:

• To review state-of-the-art 3-D wavelet coders for volumetric medical data.
• To propose novel compression algorithms addressing the shortcomings of current DCT-based approaches.
• To benchmark new methods against existing 3-D DCT techniques.

Main Methods:

• Overview of advanced 3-D wavelet coding techniques.
• Development of new compression methods utilizing quadtree and block-based coding, layered zero-coding, and context-based arithmetic coding.
• Implementation of a new 3-D DCT-based scheme for comparative analysis.

Main Results:

• Proposed wavelet-based algorithms generate embedded data streams supporting lossless decoding and scalability.
• Algorithms meet essential functionality constraints for medical applications.
• Objective and subjective evaluations demonstrate competitive lossy and lossless compression performance compared to state-of-the-art methods.

Conclusions:

• The proposed wavelet-based algorithms effectively address the limitations of 3-D DCT for medical volumetric data coding.
• These methods provide a scalable and lossless compression solution suitable for medical applications.
• The new algorithms offer competitive compression efficiency for both lossy and lossless scenarios.