Advanced Severity Detection in Histopathological Ovarian Cancer: Rank-Based Leaf Wind Optimization and Alpha...

Area of Science:

• Medical Imaging
• Computational Pathology
• Artificial Intelligence in Oncology

Background:

• Ovarian cancer (OC) frequently presents in advanced stages due to subtle early symptoms.
• Accurate staging is critical for effective ovarian cancer treatment and patient outcomes.

Purpose of the Study:

• To develop and validate a novel methodology for assessing ovarian cancer severity using histopathological image analysis.
• To enhance the accuracy and efficiency of ovarian cancer diagnosis and staging.

Main Methods:

• Image preprocessing included normalization and Contrast Limited Adaptive Histogram Equalization.
• Feature extraction utilized ResNet 50 and Inception v4 architectures.
• A Rank-Based Leaf in Wind Optimization and Alpha Piecewise Linear Fuzzy approach was employed for severity classification.

Main Results:

• The proposed model achieved 99.25% accuracy and 97.98% precision in classifying tumor severity.
• The methodology demonstrated effectiveness in classifying severity levels even with diagnostic uncertainty.
• The system successfully prioritized key features for accurate diagnosis.

Conclusions:

• The developed methodology significantly enhances diagnostic accuracy for ovarian cancer.
• This approach supports earlier detection and more precise treatment planning.
• Future research will focus on clinical integration and model optimization.