Skip to main content

Table 1 False and right percentages of canal/sinus/fossa detections of the AI system

From: A deep learning approach for dental implant planning in cone-beam computed tomography images

 Regions Canal/sinus/fossa detection (%, n) Missing tooth detection (%, n)
Canal detection (Mandibula) Sinus/fossa detection (Maxilla) Mandibula Maxilla
False Right False Right False Right False Right
Anterior 83%
(n = 44)
17%
(n = 9)
91.8%
(n = 56)
8.2%
(n = 5)
9.4%
(n = 5)
90.6%
(n = 48)
8.2%
(n = 5)
91.8%
(n = 56)
Premolar 29.5%
(n = 18)
70.5%
(n = 43)
27.7%
(n = 28)
72.3%
(n = 73)
3.3%
(n = 2)
96.7%
(n = 59)
5.9%
(n = 6)
94.1%
(n = 95)
Molar 2.5%
(n = 3)
97.5%
(n = 117)
7.1%
(n = 8)
92.9%
(n = 104)
1.7%
(n = 2)
98.3%
(n = 118)
3.6%
(n = 4)
96.4%
(n = 108)
Total 27.8%
(n = 65)
72.2%
(n = 169)
33.6%
(n = 92)
66.4%
(n = 182)
3.8%
(n = 9)
96.2%
(n = 225)
5.5%
(n = 15)
94.5%
(n = 259)