Diagnosis of COVID-19 patients by adapting hyper parametertuned deep belief network using hosted cuckoo optimization...

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• Accurate and rapid identification of COVID-19 is crucial for effective disease management.
• Chest X-ray imaging is a common diagnostic tool for respiratory illnesses.
• Machine learning offers potential for automated analysis of medical images.

Purpose of the Study:

• To develop and evaluate a machine learning model for classifying chest X-ray images as COVID-19 positive or negative.
• To enhance a deep belief network (DBN) model using a hosted cuckoo optimization algorithm (HCOA) for hyperparameter tuning.
• To compare the proposed model's performance against existing methods like CNN-SMO and SVM-BOA.

Main Methods:

• Pre-processing of chest X-ray images to remove noise.
• Application of a deep belief network (DBN) optimized with a hosted cuckoo optimization algorithm (HCOA) for hyperparameter tuning.
• Performance evaluation using metrics such as accuracy, precision, and F-score.

Main Results:

• The proposed DBN-HCOA model achieved significant improvements in accuracy, precision, and F-score compared to baseline methods.
• Specific performance gains included higher accuracy for Normal (23.5%-28.3%) and COVID-19 (31.5%-32.3%) cases.
• The model also demonstrated enhanced precision and F-score for both Normal and COVID-19 classifications.

Conclusions:

• The hyperparameter-tuned deep belief network using hosted cuckoo optimization algorithm provides an effective method for COVID-19 detection from chest X-rays.
• This AI-driven approach shows superior performance over traditional machine learning techniques for this diagnostic task.
• The findings suggest the potential of this methodology for aiding in the rapid and accurate identification of COVID-19.