Signal analysis and compression performance evaluation of pathological microscopic images

Area of Science:

• Digital Pathology
• Medical Imaging Technology

Background:

• Telepathology relies on high-quality digitized microscopic images for remote diagnosis.
• Efficient image compression is crucial for transmitting large pathological datasets.

Purpose of the Study:

• To evaluate the digitization of pathological images at 2k x 2k resolution.
• To analyze the signal characteristics and compression performance of these images using a super-high-definition imaging system.
• To assess the impact of JPEG compression and chrominance subsampling on pathological image quality.

Main Methods:

• Direct digitization (non-film process) of diagnostic-grade pathological samples at 2k x 2k resolution.
• Analysis of color distribution and spatial frequency characteristics, comparing pathological images to general images.
• Evaluation of color differences in L*a*b* space and JPEG compression ratios.
• Subjective assessment of chrominance subsampling effects on image fidelity.

Main Results:

• Pathological images exhibit distinct characteristics, notably high spatial frequency in their chrominance components, differing from general images.
• JPEG compression performance was evaluated, with analysis of color differences in L*a*b* space.
• Subjective evaluations indicated the influence of chrominance subsampling on perceived image quality.

Conclusions:

• Understanding the unique spatial frequency characteristics of pathological images is essential for optimizing digitization and compression strategies in telepathology.
• JPEG compression, particularly chrominance subsampling, requires careful consideration to maintain diagnostic accuracy in telepathic applications.