Skip to main content

Table 2 Bone height and thickness measurements with the AI system and manual assessment

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

 Regions

Parameters

Bone height measurements (n = 428)#

Bone thickness measurements (n = 493)#

 AI

Manual

 AI

Manual

Median (min–max)

Median (min–max)

Median (min–max)

Median (min–max)

Mandibula

Anterior

19.2 (3.1–31.6)

12.8 (5.7–19.1)

5.7 (1.2–11.8)

4 (2.1–10.8)

Premolar

12 (1.2–26.4)*

12 (5–19.9)*

6.1 (2.4–11.7)

4.6 (2.5–11.1)

Molar

10 (1.2–19.2)

11,7 (2.6–20)

7.7 (4.5–13.8)

5.2 (3.3–12.5)

Maxilla

Anterior

24.3 (3.9–27.3)

13 (3.6—16,6)

6.5 (3.9–31.1)

4.4 (2.1–18.9)

Premolar

12.2 (2.3–28.8)*

12.1 (2.4—22)*

7.2 (2–34.4)

4.9 (0–11.1)

Molar

6.9 (0.5–26.3)*

7.6 (0–17.6)*

9.5 (1.6–29.5)

5.6 (0–15.2)

  1. Bold variables in the table indicates no statistical significant difference (p > 0.05)
  2. *There were no differences between AI and manual measurements for each parameter (p > 0.05). Min, minimum; Max, Maximum
  3. #A total of 508 measurements were performed, but the AI system was unable to perform 80 of bone height measurements and 15 of bone thickness measurements. For each parameter, statistical analyzes were made on the measurements that can only be evaluated with the AI system