Abstract :

Pancreatic Cancer Continues To Pose Significant Diagnostic Challenges Due To Its Late-stage Symptom Onset And Subtle Imaging Characteristics. In This Paper, A Robust Computer-aided Detection Approach Is Proposed, Integrating Deep Learning With Bio-inspired Optimization To Improve Diagnostic Accuracy Using CT Images. The Architecture Leverages The Inception V4 Network To Extract Deep Semantic Features Capable Of Capturing Minute Variations In Pancreatic Tissue. To Further Refine The Model’s Performance, Grey Wolf Optimization (GWO) Is Employed For Tuning Critical Model Parameters And Identifying The Most Relevant Features.The Inception V4 Model Identifies Key Patterns In The Images, While The Grey-Level Co-occurrence Matrix (GLCM) Extracts Important Texture Information. GWO Is Then Used To Optimize These Features For Better Classification Accuracy. The Process Includes Image Segmentation And Morphological Refinement To Highlight Areas Of Interest, Such As Possible Tumors. In Addition To Deep Learning Features, Statistical Texture Attributes Derived From The Gray-Level Co-occurrence Matrix (GLCM) Namely Contrast, Energy, And Entropy Are Computed To Characterize Fine-grained Texture Irregularities That Often Correlate With Malignancy. The Fusion Of Deep And Handcrafted Features Enhances Classification Reliability. Experimental Validation Confirms That The Combined Method Achieves Superior Performance. The Fusion Of Inception V4 And GWO Enhances Both The Sensitivity (detecting Cancer When Present) And Specificity (minimizing False Positives), Promising A Significant Improvement In Medical Image Analysis For Pancreatic Cancer Detection. Keywords— Pancreatic Cancer Detection, Deep Learning, Inception V4, Grey Wolf Optimization (GWO),Grey Level Co-occurrence Matrix (GLCM), CT Imaging.

Published:

19-11-2025

Issue:

Vol. 25 No. 11 (2025)

Page Nos:

155-159

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite