ai-corona: Radiologist-assistant deep learning framework for COVID-19 diagnosis in chest CT scans

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Deep Learning

Background:

• The COVID-19 pandemic highlighted the need for advanced diagnostic tools.
• Artificial intelligence (AI) offers potential solutions for medical image analysis.

Purpose of the Study:

• To introduce ai-corona, a deep learning framework for COVID-19 diagnosis using chest CT scans.
• To evaluate the performance of ai-corona compared to existing models and its impact on expert diagnosis.

Main Methods:

• Development of a radiologist-assistant deep learning framework named ai-corona.
• Utilizing an EfficientNetB3-based feature extractor.
• Training and evaluation on three diverse datasets (CC-CCII, MDH, MosMedData) comprising 7184 scans.

Main Results:

• ai-corona achieved high AUC scores: 0.997 (CC-CCII), 0.989 (MDH), and 0.954 (MosMedData).
• The framework outperformed alternative models in COVID-19 diagnosis.
• Incorporating ai-corona into expert workflows increased diagnostic speed and accuracy.

Conclusions:

• ai-corona is an effective deep learning tool for COVID-19 detection in chest CT scans.
• The framework demonstrates superior performance and aids in enhancing radiologist efficiency.
• AI-assisted diagnosis holds significant promise for future medical systems and pandemic response.