highly significantp < 0.001
When comparing the overall binary AI outputs (presence/absence of abnormality), McNemar’s test revealed a highly significant difference between OPG and CBCT (χ 2 = 16.06, p < 0.001).
When comparing the overall binary AI outputs (presence/absence of abnormality), McNemar’s test revealed a highly significant difference between OPG and CBCT (χ 2 = 16.06, p < 0.001).