Markov texture parameters as prognostic indicators in endometrial cancer

Area of Science:

• Oncology
• Medical Imaging
• Biostatistics

Background:

• Optical texture is a surface property distinct from color and shape.
• Image analysis enables quantitative measurement of optical texture in grayscale images.
• Endometrial cancer recurrence prediction remains a critical clinical challenge.

Purpose of the Study:

• To investigate the prognostic relevance of Markov texture parameters in endometrial cancer.
• To determine if image analysis of optical texture can predict disease recurrence.
• To identify novel independent prognostic indicators for endometrial cancer.

Main Methods:

• Prospective study of 74 surgically treated endometrial cancer patients.
• Quantification of 21 Markov texture parameters and DNA index (DI) using image analysis.
• Evaluation of traditional prognostic factors including FIGO stage, grade, and depth of invasion.

Main Results:

• 14 Markov parameters correlated significantly with patient survival (P < 0.05).
• Eleven Markov parameters showed significant correlation with increasing FIGO stage (P < 0.05).
• Difference entropy, information measure B, and diagonal moment emerged as independent prognostic indicators.

Conclusions:

• Image analysis effectively quantifies optical texture for prognostic assessment.
• Three Markov texture parameters (difference entropy, information measure B, diagonal moment) are independent predictors of outcome in endometrial cancer.
• These texture parameters offer potential as novel biomarkers alongside established prognostic factors.