Skip to main content
. 2023 Apr 4;27(5):2255–2265. doi: 10.1007/s00784-023-04978-4

Table 2.

Results accuracy artificial intelligence versus experts

Results accuracy
AI Experts E1 E2 E3
Combined error (mm) 2.73 2.79 2.79 2.79 2.78
Combined error (%) 100% +2.1% +2.4% +2.3% +1.7%
Z=3145 p=0.71 Z=3244 p=0.92 Z=3146 p=0.71 Z=3227 p=0.88
Combined normal error (%) 100% + 0.8% +0.6% +0.8% +1%
Z=3234 p=0.9 Z=3127 p=0.67 Z=3225 p=0.88 Z=3244 p=0.92
Median points (7 points) (mm) 1.75 1.5*** 1.47*** 1.51*** 1.52**
Z=2103 Z=2035 Z=2111 Z=2273
Osteotomy points (12 points) (mm) 3.27 2.93** 2.98** 2.93* 2.89**
Z=2340 Z=2331 Z=2398 Z=2332
Other points (16 points) (mm) 2.75 3.24*** 3.23*** 3.25*** 3.24***
Z=1702 Z=1763 Z=1760 Z=1711

AI, artificial intelligence; E1, expert 1; E2, expert 2; E3, expert 3: combined error in mm and % as well as the combined normal error in % of the artificial intelligence and the three experts for all landmarks together as well as divided into three categories (median points, osteotomy points, other points). Statistically significant differences are marked with ***p < 0.001; **p < 0.01; *p < 0.05