Deep Learning-Based Computer-Aided Diagnosis (CAD): Applications for Medical Image Datasets.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Machine Learning

Background:

• Computer-aided diagnosis (CAD) systems are crucial for accurate medical predictions.
• Deep learning methods have shown significant promise in analyzing medical image datasets.

Purpose of the Study:

• To develop an automated CAD system for highly accurate medical diagnoses.
• To integrate deep learning, meta-heuristic, and supervised machine learning algorithms for enhanced diagnostic performance.

Main Methods:

• Utilized pre-trained convolutional neural networks (CNNs) or auto-encoders for feature extraction.
• Employed Ant Colony Optimization (ACO) for efficient feature selection, reducing data dimensionality.
• Implemented learnable classifiers for final diagnosis prediction (classification).

Main Results:

• The proposed framework achieved 99.61% accuracy in diagnosing COVID-19 from chest X-rays (CXR).
• The system demonstrated 99.18% accuracy in detecting brain tumors from magnetic resonance imaging (MRI).
• The approach outperformed existing state-of-the-art methods on both datasets.

Conclusions:

• The developed CAD system offers a reliable and accurate tool for physicians and radiologists.
• The novel framework combining deep learning, auto-encoders, and ACO shows high potential for clinical diagnostic applications.
• The system can be confidently used for diagnosing COVID-19 and specific brain tumors.