Deep learning aided decision support for pulmonary nodules diagnosing: a review

Area of Science:

• Medical Imaging Analysis
• Artificial Intelligence in Healthcare

Background:

• Computer-assisted diagnosis for pulmonary nodules has a long research history.
• Deep learning shows significant progress in analyzing medical images for clinical diagnosis.

Purpose of the Study:

• To provide a comprehensive review of deep learning techniques for pulmonary nodule diagnosis.
• To highlight deep learning as a decision support tool for analyzing chest scan data.

Main Methods:

• Review of state-of-the-art deep learning methodologies.
• Analysis of applications in nodule detection, feature extraction, and classification.
• Focus on deep learning for pulmonary nodule diagnosis.

Main Results:

• Deep learning techniques offer promising solutions for pulmonary nodule diagnosis.
• These methods effectively address challenges like feature extraction and classification.
• Significant advancements have been made in both academia and industry.

Conclusions:

• Deep learning is a powerful alternative for decision support in pulmonary nodule diagnosis.
• This review is the first to exclusively focus on deep learning for this application.
• Future research can leverage these findings for improved diagnostic accuracy.