Effective Utilization of Multiple Convolutional Neural Networks for Chest X-Ray Classification

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computer Science

Background:

• Radiography is a vital, affordable medical imaging technique for disease diagnosis.
• Chest Radiography is crucial but interpretation challenges exist due to radiologist scarcity and human error.
• Artificial Intelligence (AI) and Convolutional Neural Networks (CNNs) are explored to enhance diagnostic accuracy.

Purpose of the Study:

• To present a novel method for parallelizing multiple CNN architectures for Chest X-ray classification.
• To evaluate the performance of parallelized CNNs against existing architectures.
• To assess the effectiveness of the proposed AI approach in improving diagnostic accuracy.

Main Methods:

• Developed a novel parallelization technique for multiple CNN architectures.
• Applied the method to Chest X-ray classification tasks.
• Conducted comprehensive evaluations using four large-scale datasets, including a non-medical dataset.

Main Results:

• Achieved improved accuracy in Chest X-ray classification.
• Demonstrated superior performance for 9 out of 13 and 11 out of 14 labels on key evaluation datasets.
• Validated the effectiveness of parallelized CNNs in medical image analysis.

Conclusions:

• The proposed parallelization method shows significant potential for enhancing AI-driven medical image diagnostics.
• The study highlights the benefits of combining multiple CNN architectures for improved accuracy.
• Future work will focus on system limitations and further improvements.