Incremental benefit of computer-aided detection when used as a second and concurrent reader of CT colonographic data:...

Area of Science:

• Radiology
• Medical Imaging
• Gastroenterology

Background:

• Computed tomographic (CT) colonography is a key tool for colorectal cancer screening.
• The effectiveness of CT colonography relies on accurate polyp detection by radiologists.
• Computer-aided detection (CAD) systems aim to improve diagnostic accuracy.

Purpose of the Study:

• To evaluate the impact of computer-aided detection (CAD) on radiologist performance in interpreting CT colonography data.
• To compare the effectiveness of CAD as a second reader versus a concurrent reader.

Main Methods:

• 16 experienced radiologists interpreted CT colonography cases of 112 patients (56 with polyps).
• Each case was reviewed under three conditions: unassisted, second-read CAD, and concurrent CAD.
• Reader performance (sensitivity and specificity) was compared against reference-truth interpretation.

Main Results:

• Second-read CAD significantly increased per-patient sensitivity for polyp detection (7.0% increase).
• Concurrent CAD also improved per-patient sensitivity (4.5% increase), but less effectively.
• Second-read CAD significantly improved per-polyp sensitivity for both large and small polyps, while concurrent CAD only improved detection of small polyps.

Conclusions:

• Utilizing CAD as a second reader substantially enhances radiologist detection of polyps in CT colonography.
• Concurrent CAD application is less beneficial compared to a second-read approach for improving detection rates.