Exploration of Medicine

From: Artificial intelligence in panoramic images—clinical aid to a dentist

Table 5

Showing inter-rater reliability for caries

Cohen’s weighted kappa
Ratings	Weighted kappa^a	Asymptotic			95% Asymptotic confidence interval
Ratings	Weighted kappa^a	Standard error^b	z^c	Sig.	Lower bound	Upper bound
observer 1–observer 2	0.844	0.016	24.649	< 0.001	0.814	0.875
observer 1–observer 3	0.784	0.019	22.818	< 0.001	0.747	0.821
observer 1–observer 4	0.838	0.017	24.389	< 0.001	0.804	0.872
observer 1–AI	0.903	0.013	26.418	< 0.001	0.877	0.928
observer 2–observer 3	0.789	0.020	22.862	< 0.001	0.750	0.827
observer 2–observer 4	0.821	0.019	23.726	< 0.001	0.784	0.858
observer 2–AI	0.903	0.013	26.217	< 0.001	0.877	0.928
observer 3–observer 4	0.787	0.019	22.617	< 0.001	0.748	0.825
observer 3–AI	0.845	0.018	24.510	< 0.001	0.808	0.881
observer 4–AI	0.884	0.016	25.381	< 0.001	0.853	0.916

Return to article view

^a The estimation of the weighted kappa uses linear weights; ^b value does not depend on either null or alternative hypotheses; ^c estimates the asymptotic standard error assuming the null hypothesis that weighted kappa is zero. Sig: significance