Context adaptive lossless and near-lossless coding for digital angiographies.

Area of Science:

• Biomedical Imaging
• Image Processing
• Data Compression

Background:

• Hemodynamic image sequences are crucial for medical diagnostics.
• Existing compression methods struggle with noise and intensity variations in these images.
• Lossless and near-lossless compression is vital for preserving diagnostic information.

Purpose of the Study:

• To develop a context adaptive coding method for hemodynamic image sequences.
• To enhance compression efficiency while maintaining high image quality.
• To improve robustness against noise and intensity variations.

Main Methods:

• Implementation of a two-stage context adaptive linear predictor for motion compensation.
• Application of the coding method to 12 bits/pixel hemodynamic image data.
• Comparative analysis against JPEG-2000, JPEG-LS, and a prior method [1].

Main Results:

• The proposed method demonstrated superior lossless compression performance.
• Achieved 3.8%, 2%, and 1.6% better compression than JPEG-2000, JPEG-LS, and method [1], respectively.
• Performance gains are expected to increase with near-lossless compression.

Conclusions:

• The context adaptive coding method offers significant improvements for hemodynamic image compression.
• The method is robust to common image degradations like noise and intensity changes.
• This approach holds potential for advancing medical image archiving and transmission.