Table 2 Univariate and multiple logistic regression analyses.

From: Using deep learning to predict temporomandibular joint disc perforation based on magnetic resonance imaging

Variables

Univariate

Multiple

OR (95% CI)

p-value

OR (95% CI)

p-value

Age

1.06 (1.04–1.09)

0.000***

1.05 (1.02–1.09)

0.002**

Sex

Male

1

 

1

 

Female

2.37 (1.17–5.14)

0.021*

2.04 (0.76–5.83)

0.276

Shape of the disc

Biconcave

1

 

1

 

Folded

4.57 (0.76–87.59)

0.165

1.12 (0.13–24.74)

0.892

Flattened

15.00 (2.82–278.31)

0.011*

6.27 (0.98–123.82)

0.075

Eyeglasses-shaped

26.25 (5.19–479.44)

0.002**

11.72 (1.55–251.4)

0.040*

Amorphous

118.33 (22.97–2178.49)

0.000***

44.92 (5.84–975.71)

0.002**

Signal intensity of bone marrow

Normal

1

 

1

 

Low

4.80 (2.61–9.25)

0.000***

3.47 (1.48–8.56)

0.005**

Fluid collection

G0

1

   

G1

1.36 (0.82–2.27)

0.234

  

G2

1.14 (0.56–2.29)

0.716

  

G3

2.66 (0.76—10.55)

0.133

  

Disc displacement

Normal + ADcR

1

   

Early ADsR

4.03 (1.71–10.69)

0.002**

1.45 (0.45–4.93)

0.367

Late ADsR

11.11 (5.01–28.28)

0.000***

0.94 (0.25–3.63)

0.898

Joint space

Normal

1

 

1

 

Narrowing

3.95 (1.79–10.02)

0.002**

3.21 (1.09–10.74)

0.066

Bone to bone contact (close)

21.71 (5.39–115.90)

0.000***

9.19 (1.51–70.15)

0.047*

Bone to bone contact (open)

57 (13.09–413.06)

0.000***

45.4 (6.53–497.03)

0.000***

Changes of condyle and fossa

2 or less

1

 

1

 

2 or more

6.23 (3.51–11.47)

0.000***

3.79 (1.75–8.58)

0.002**

  1. ADcR anterior displacement with reduction, ADsR anterior displacement without reduction.
  2. *p < 0.05, **p < 0.01, ***p < 0.001.