Coefficient-Shuffled Variable Block Compressed Sensing for Medical Image Compression in Telemedicine Systems.

Area of Science:

• Medical Imaging
• Signal Processing
• Data Compression

Background:

• Medical imaging is vital for diagnosing conditions, but generates large datasets requiring compression for efficient analysis and transmission.
• Existing Compressed Sensing (CS) methods, including block-based CS (BCS), face challenges in high-quality image reconstruction due to random sampling.
• Effective compression is essential for managing large volumes of medical data, especially in long-term patient monitoring.

Purpose of the Study:

• To introduce a novel Compressed Sensing (CS) method, coefficient shuffling variable BCS (CSEM-VBCS), for medical image compression.
• To improve image reconstruction quality and achieve higher compression ratios compared to existing techniques.
• To address the limitations of random sampling in conventional CS and BCS methods.

Main Methods:

• Developed a novel CS method, CSEM-VBCS, utilizing an energy matrix and coefficient shuffling.
• Applied the CSEM-VBCS method to compress diverse medical images with balanced sparsity.
• Evaluated performance metrics against contemporary state-of-the-art compression techniques.

Main Results:

• The proposed CSEM-VBCS method demonstrated a substantial compression ratio and good reconstruction quality.
• Experimental evaluations showed remarkable enhancement in performance metrics compared to existing methods.
• CSEM-VBCS effectively prioritizes regions of interest through coefficient shuffling, improving compression without compromising image quality.

Conclusions:

• CSEM-VBCS offers a superior approach to medical image compression, balancing compression ratio and reconstruction quality.
• The method is particularly beneficial for telemedicine applications, overcoming bandwidth limitations for high-resolution medical image transmission.
• CSEM-VBCS enhances the efficiency of remote patient monitoring and diagnosis through faster data acquisition and reduced redundancy.