Fifty years of computer analysis in chest imaging: rule-based, machine learning, deep learning.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Pulmonology

Background:

• Computer-aided diagnosis (CAD) has been utilized in medical imaging for 50 years.
• Pulmonary imaging, including radiography and CT, is a key area for CAD applications.
• Traditional rule-based approaches have been the standard for CAD in lung imaging.

Purpose of the Study:

• To trace the evolution of CAD technologies in pulmonary imaging.
• To compare the efficacy of rule-based, machine learning, and deep learning approaches in lung CAD.
• To highlight the advancements and current trends in deep learning for chest imaging analysis.

Main Methods:

• Review of historical development in computer-aided diagnosis for lung imaging.
• Comparative analysis of rule-based, machine learning, and deep learning algorithms.
• Illustration of differences through various CAD applications in chest imaging.

Main Results:

• Machine learning surpassed classical rule-based methods in lung CAD.
• Deep learning demonstrates even greater performance improvements over machine learning.
• Significant advancements in CAD for pulmonary diagnostics are driven by deep learning.

Conclusions:

• Deep learning represents the current state-of-the-art in computer-aided diagnosis for pulmonary imaging.
• The progression from rule-based systems to deep learning has significantly enhanced CAD capabilities in the chest.
• Future research in lung imaging CAD will likely focus on further leveraging deep learning.