Lung nodule computer-aided detection as a second reader: influence on radiology residents

Area of Science:

• Radiology
• Medical Imaging
• Artificial Intelligence in Medicine

Background:

• Pulmonary nodules require accurate detection for timely diagnosis and treatment.
• Radiology residents are in training and may benefit from assistive tools.
• Computer-aided detection (CAD) software is increasingly used in medical imaging.

Purpose of the Study:

• To assess the efficacy of CAD software as a supplementary tool for radiology residents.
• To evaluate if CAD improves pulmonary nodule detection rates among residents.

Main Methods:

• 110 computed tomography (CT) cases with 4-30 mm nodules were used.
• Radiology residents interpreted cases with and without CAD software.
• Performance was compared against a ground truth established by expert radiologists.

Main Results:

• Residents' mean sensitivity increased from 85% without CAD to 90% with CAD (P < 0.05).
• The CAD-assisted resident performance became comparable to that of board-certified radiologists (89% mean sensitivity).

Conclusions:

• CAD software serves as an effective second reader for enhancing resident performance.
• Utilizing CAD can bridge the diagnostic accuracy gap between residents and experienced radiologists in pulmonary nodule detection.