Interframe coding of magnetic resonance images

Area of Science:

• Medical Imaging
• Image Compression
• Signal Processing

Background:

• Interframe coding for medical images, especially magnetic resonance (MR) images, has faced challenges due to unsuitable estimation models and inherent MR noise.
• Previous attempts to leverage interframe redundancies in MR image coding have been largely unsuccessful.

Purpose of the Study:

• To develop a novel interframe coding method for MR images that addresses the limitations of existing techniques.
• To improve the efficiency and effectiveness of MR image compression for medical applications.

Main Methods:

• A continuous affine mapping model, optimized by deforming triangles, is employed for interframe estimation.
• A median filter is integrated within the estimation loop to mitigate the effects of inherent magnetic resonance imaging (MRI) noise.
• Residue frames are quantized using a zero-tree wavelet coder with arithmetic entropy coding.

Main Results:

• The proposed method successfully utilizes interframe redundancies in MR images, a feat not achieved by prior techniques.
• The integration of a median filter effectively handles the inherent noise in MRI data.
• The zero-tree wavelet coder with arithmetic entropy coding enables progressive transmission of quantized residue frames.

Conclusions:

• The developed interframe coding method offers a significant advancement in MR image compression.
• The technique's ability to handle noise and support progressive transmission makes it highly attractive for medical imaging applications.
• This approach enhances the efficiency of storing and transmitting medical images, potentially improving diagnostic workflows.