Table 5.

Results of the agreement analyses

Lengths and angles	Sample size (N)	ICC between AI and GT [95% CI]	Intra-reader reliability [95% CI]	ICC between radiologists [95% CI]
HKA	331	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]	> 0.99 [0.99, 1]
Pelvic obliquity	150	0.97 [0.96, 0.98]	0.98 [0.95, 0.99]	0.99 [0.98, 0.99]
Top leg length	309	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]	> 0.99 [0.95, 1]
Center leg length	331	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]
Top femoral length	312	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]
Center femoral length	334	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]
Tibial length	344	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]	> 0.99 [> 0.99, 1]

Agreement between AI and the ground truth (GT) was assessed with intraclass correlation coefficients (ICC) from a two-way mixed-effects model with absolute agreement for multiple raters. The intra-reader reliability was assessed with ICC from a two-way mixed-effects model with absolute agreement for a single rater. Agreement between the two radiologists who established the ground truth was evaluated with ICC from a two-way random-effects model with absolute agreement for multiple raters. Sample size (N) is also displayed