Classifying chest CT images as COVID-19 positive/negative using a convolutional neural network ensemble model and...

Area of Science:

• Medical Imaging Analysis
• Artificial Intelligence in Healthcare
• Radiology

Background:

• Accurate and rapid classification of chest CT images is crucial for diagnosing coronavirus disease 2019 (COVID-19).
• Developing effective AI models for medical image analysis is an ongoing research area.

Purpose of the Study:

• To develop and evaluate a novel ensemble model for classifying chest CT images as positive or negative for COVID-19.
• To improve the accuracy and efficiency of COVID-19 detection using medical imaging.

Main Methods:

• A convolutional neural network (CNN) ensemble model, termed COVID19-CNN, was developed.
• The model utilizes a majority voting strategy to combine predictions from multiple trained CNN models.
• CNN models were trained using transfer learning from pre-trained models, with hyperparameters optimized via uniform experimental design.

Main Results:

• The COVID19-CNN ensemble model achieved 96.7% accuracy in classifying chest CT images for COVID-19.
• Performance metrics including precision, recall, specificity, and F1-score surpassed those of individual CNN models.
• The model demonstrated superior performance compared to individual CNN models in COVID-19 detection.

Conclusions:

• The COVID19-CNN ensemble model exhibits excellent capability and superior accuracy for classifying chest CT images.
• This AI-driven approach shows promise for efficient and reliable COVID-19 diagnosis.