Integrating Ant Colony Optimization with Deep Learning for Improved Lung Cancer Diagnosis and Prognosis

Area of Science:

• Artificial Intelligence in Medicine
• Computational Intelligence
• Medical Imaging Analysis

Background:

• Lung cancer diagnosis relies on accurate interpretation of CT images and medical records.
• Current diagnostic methods can be enhanced by advanced computational techniques.
• AI-driven healthcare solutions show promise for improving diagnostic precision.

Purpose of the Study:

• To improve the accuracy and reliability of lung cancer diagnosis.
• To integrate Ant Colony Optimization (ACO) with deep learning models for enhanced detection.
• To establish a foundation for future AI-driven healthcare solutions in oncology.

Main Methods:

• Ensemble learning combining Ant Colony Optimization (ACO) with deep learning models: DenseNet, ResNet 50, VGG 19, and Long Short-Term Memory (LSTM) networks.
• ACO was employed for feature selection to optimize model performance.
• The integrated models were applied to diagnose lung cancer from CT images and medical records.

Main Results:

• The ensemble model integrating ACO with deep learning significantly improved lung cancer diagnostic accuracy.
• DenseNet combined with ACO and LSTM achieved the highest accuracy at 97.9%.
• ResNet 50 reached 96.2% accuracy, and VGG 19 achieved 92.3% accuracy, demonstrating the effectiveness of ACO in enhancing model performance.

Conclusions:

• Ant Colony Optimization effectively optimizes feature selection, leading to significant improvements in deep learning model performance for lung cancer diagnosis.
• The study demonstrates the potential of combining swarm intelligence with deep learning for more precise and reliable medical diagnoses.
• This approach offers a promising advancement for AI-driven healthcare, improving lung cancer diagnosis and patient outcomes.