Gray-Level Co-occurrence Matrix Analysis of Nuclear Textural Patterns in Laryngeal Squamous Cell Carcinoma: Focus on...

Area of Science:

• Computational pathology
• Digital image analysis
• Cancer diagnostics

Background:

• Gray-level co-occurrence matrix (GLCM) and discrete wavelet transform (DWT) are advanced computational techniques for analyzing tissue texture.
• These methods show promise in cancer identification and classification within pathology.
• Morphologically intact squamous epithelial cells in laryngeal carcinoma may possess unique textural characteristics.

Purpose of the Study:

• To investigate distinct nuclear textural features of squamous epithelial cells in laryngeal carcinoma using GLCM and DWT.
• To develop and evaluate machine learning models for classifying laryngeal carcinoma cells based on these textural features.

Main Methods:

• Analysis of nuclear textural features using Gray-level co-occurrence matrix (GLCM) and discrete wavelet transform (DWT).
• Comparison of GLCM indicators (angular second moment, inverse difference moment, textural contrast) between cancerous and noncancerous laryngeal tissues.
• Development of machine learning models (random forests, support vector machine) utilizing GLCM and DWT quantifiers as input.

Main Results:

• Significant differences in average nuclear GLCM indicators were observed between laryngeal carcinoma cells and noncancerous cells.
• Machine learning models demonstrated good classification accuracy and discriminatory power in separating cell types.
• The models were successfully trained using GLCM and DWT quantifiers on a limited cell sample.

Conclusions:

• Squamous epithelial cells in laryngeal carcinoma exhibit unique nuclear textural features detectable by GLCM and DWT.
• Machine learning models based on these textural features show potential for accurate cancer cell classification.
• These findings support the development of AI-based sensors for laryngeal carcinoma diagnostic protocols.