Fig. 5: System’s performance when suing different feature extracting. | Nature Communications

Fig. 5: System’s performance when suing different feature extracting.

From: AI-assisted cervical cytology precancerous screening for high-risk population in resource-limited regions using a compact microscope

Fig. 5

The ROC curves of Supervised, Pretrained and Pretrained on ImageNet (ImageNet for short in legend) features on InTest, NewCohorts, External-A and External-B sets. AUC values and sample sizes are added in the legend and above the figure, respectively. Source data are provided as a Source Data file.

Back to article page