[A nucleus area extraction method for image analysis of kidney-tissue slice].

Area of Science:

• Nephrology
• Biomedical Imaging
• Computational Pathology

Context:

• Accurate analysis of human kidney glomeruli nuclei is crucial for disease diagnosis.
• Slice preparation artifacts can lead to miss or incorrect nuclei extraction, hindering analysis.
• Existing segmentation methods struggle with nuclei conglutination and extraction errors.

Purpose:

• To develop a robust and precise method for segmenting nuclei in human kidney glomeruli.
• To overcome challenges associated with nuclei extraction and conglutination in histological slices.
• To improve the accuracy of quantitative analysis of kidney glomeruli.

Summary:

• A dynamic thresholding method utilizing an eigenvalue feedback strategy is proposed for nuclei segmentation in human kidney glomeruli.
• The method leverages R channel (RGB) and C channel (CMYK) information, combined with Laplacian of Gaussian (LOG) filtering, to identify nuclei positions.
• Eigenvalue feedback adjusts the threshold surface to resolve nuclei conglutination, enabling accurate segmentation by integrating positional information.

Impact:

• The proposed method demonstrates high precision in nuclei segmentation, facilitating more reliable quantitative analysis of kidney pathology.
• This technique has the potential to enhance diagnostic accuracy in nephrology and kidney disease research.
• Improved nuclei segmentation can aid in the development of automated diagnostic tools for renal pathologies.