Skip to main content
Fig. 3 | BMC Medicine

Fig. 3

From: Development and validation of an artificial intelligence system for grading colposcopic impressions and guiding biopsies

Fig. 3

The ROC curve of diagnostic performance for detecting pathological HSIL+ at different hypothetical biopsy thresholds. Note: The hypothetical biopsy thresholds are shown: A) normal/benign versus low-grade or worse and B) less severe impressions versus high-grade or worse. a Analysis of images alone by CAIADS. b Analysis of both images and non-image information by CAIADS. c Analysis of both images and non-image information by colposcopists. Abbreviations: CAIADS, Colposcopic Artificial Intelligence Auxiliary Diagnostic System; AUC, area under the curve. Less severe impressions included normal/benign and low-grade. Non-image information included the patient’s primary screening findings (cytology, HPV status)

Back to article page

BMC Medicine

ISSN: 1741-7015

Contact us