Advanced computer-aided detection system exhibits no more false positives than experienced endoscopists in an image-based comparative study of colon polyps - Report - MDSpire
Advertisement
Advanced computer-aided detection system exhibits no more false positives than experienced endoscopists in an image-based comparative study of colon polyps
Clinical Report: Comparative Study of CADe and Endoscopists in Colonoscopy
Overview
This study compares the false positive rates and diagnostic performance of a computer-aided detection (CADe) system and experienced endoscopists in identifying colon polyps. Results indicate that the CADe system exhibits a similar number of false positives as skilled endoscopists while maintaining high sensitivity for polyp detection.
Background
Colorectal cancer is a leading cause of cancer-related mortality worldwide, making effective polyp detection critical. Colonoscopy is the primary method for identifying and removing precancerous polyps, yet challenges remain in visualizing and accurately diagnosing these lesions. The introduction of CADe systems aims to enhance detection rates, but concerns about false positives could impact clinician trust and procedural efficiency.
Data Highlights
Parameter
CADe
Endoscopists
P-value
Sensitivity
90.09%
87.56%
0.46
Specificity
86.64%
85.41%
0.79
Accuracy
88.36%
86.48%
0.43
Key Findings
The CADe system demonstrated a sensitivity of 90.09% compared to 87.56% for endoscopists (P = 0.46).
Specificity was 86.64% for CADe and 85.41% for endoscopists (P = 0.79).
Overall accuracy was 88.36% for CADe versus 86.48% for endoscopists (P = 0.43).
In fully inflated lumen images, CADe sensitivity was 92.24% compared to 92.61% for endoscopists (P = 1.00).
In deflated lumen images, CADe sensitivity was 87.93%, while endoscopists had 82.51% (P = 0.35).
False positives from CADe were comparable to those from experienced endoscopists.
Clinical Implications
Understanding the performance of CADe systems in relation to experienced endoscopists can help inform their integration into clinical practice. The similar false positive rates suggest that CADe may be a viable tool for enhancing polyp detection without compromising diagnostic accuracy.
Conclusion
The study indicates that a state-of-the-art CADe system performs comparably to experienced endoscopists in terms of false positives and sensitivity for polyp detection.
