Figure 4.
Performance comparison between the artificial intelligence algorithm and two surgeons. (A) Three-way analysis between three groups; (B) Pairwise analysis showed significantly lower accuracy in the RUL of surgeon A compared to the AI algorithm; (C) subgroup analysis between non-contrast vs. contrast CT and lobectomy vs. segmentectomy. NECT, non-enhanced CT; CECT, contrast-enhanced CT. *p<0.05; ns, non-significant.