A Novel Weighted Consensus Machine Learning Model for COVID-19 Infection Classification Using CT Scan Images

Area of Science:

• Medical Imaging Analysis
• Artificial Intelligence in Healthcare
• Infectious Disease Diagnostics

Background:

• Rapid COVID-19 spread necessitates quick diagnostic methods.
• Chest X-ray and CT imaging are crucial for COVID-19 diagnosis.
• Minimizing false negatives and positives is vital for public health and patient well-being.

Purpose of the Study:

• To develop a novel weighted consensus model for COVID-19 detection.
• To minimize false negatives and false positives in diagnostic imaging.
• To maintain high accuracy while improving diagnostic reliability.

Main Methods:

• Utilized traditional Machine Learning algorithms: Linear Regression, Support Vector Machine, k-Nearest Neighbors, Decision Tree, and Random Forest.
• Developed a novel weighted consensus model incorporating normalized accuracy of individual classifiers.
• Evaluated model performance based on a predefined threshold for class prediction.

Main Results:

• Achieved high diagnostic accuracy of 99.64%, comparable to state-of-the-art techniques.
• Successfully reduced both false negatives and false positives in COVID-19 detection.
• The weighted consensus model demonstrated effective class prediction for improved diagnostic insights.

Conclusions:

• The proposed weighted consensus model effectively minimizes false negatives and positives in COVID-19 detection.
• This method offers a reliable and accurate approach to diagnosing COVID-19 using medical imaging.
• The model provides a valuable tool for enhancing diagnostic accuracy and patient management in pandemic scenarios.