Fig. 5

Performance tests and time duration comparisons between patients without and with the best-performing CNN model assistance, according to (A) GPs and (B) DSs. Statistically significant differences were assessed via paired t tests. *** p < 0.001