Computer-aided diagnosis in chest radiography: beyond nodules

Area of Science:

• Radiology
• Medical Imaging
• Computer-Aided Diagnosis

Background:

• Chest radiographs are the most frequent radiological examinations.
• They are crucial for diagnosing diseases with significant mortality and morbidity.
• Radiographs can present numerous subtle findings requiring expert interpretation.

Purpose of the Study:

• To survey areas in chest radiography beyond pulmonary nodule detection that could benefit from computer-aided detection and diagnosis (CAD).
• To identify specific applications for CAD, including interstitial infiltrates, catheter tip detection, size measurements, pneumothorax detection, and emphysema quantification.
• To discuss the current state and challenges of CAD development in these areas.

Main Methods:

• Literature review and survey of recent research in computer-aided detection and diagnosis (CAD) for chest radiography.
• Focus on applications beyond the well-established area of pulmonary nodule detection.
• Analysis of research trends and identification of under-explored areas.

Main Results:

• Several areas in chest radiography, including interstitial infiltrates, catheter tip detection, size measurements, pneumothorax, and emphysema, show potential for CAD applications.
• The volume of research dedicated to CAD in these specific areas is currently modest.
• Existing research indicates a need for further development and validation of CAD tools.

Conclusions:

• Computer-aided detection and diagnosis (CAD) has significant potential to enhance the interpretation of chest radiographs in various applications beyond pulmonary nodules.
• The current pace of research in these areas is slow, indicating a need for increased focus and investment.
• Further research and development are essential to realize the full benefits of CAD in improving diagnostic accuracy and efficiency in chest radiography.