Skip to main content

Table 1 Univariate logistic regression analysis of computed tomography features of nodular hyperplasia and follicular neoplasm

From: Comparison of computed tomography features between follicular neoplasm and nodular hyperplasia

CT features

Nodular hyperplasia (n = 65)

Follicular neoplasm (n = 59)

p value

Degree of attenuation

 no visualization

5 (7.7 %)

0 (0 %)

<0.001

 low

44 (67.7 %)

57 (96.6 %)

 

 iso-

15 (23.1 %)

1 (1.7 %)

 

 high

1 (1.5 %)

1 (1.7 %)

 

Pattern of attenuation

 homogeneous

25 (38.4 %)

19 (32.2 %)

0.466

 inhomogeneous

40 (61.6 %)

40 (67.8 %)

 

Configuration

 intraglandular

41 (63.1 %)

21 (35.6 %)

0.003

 expansile

18 (27.7 %)

34 (57.6 %)

 

 exophytic

6 (9.2 %)

4 (6.8 %)

 

Margin

 smooth

42 (64.7 %)

34 (57.6 %)

0.023

 irregular

1 (1.5 %)

0 (0 %)

 

 lobulated

16 (24.6 %)

25 (42.4 %)

 

 poorly defined

6 (9.2 %)

0 (0 %)

 

Shape

 ovoid

26 (40 %)

10 (16.9 %)

0.008

 round

29 (44.6 %)

30 (50.9 %)

 

 taller-than-wide

10 (15.4 %)

19 (32.2 %)

 

Calcifications

 none

59 (90.8 %)

48 (81.3 %)

0.186

 rim (eggshell)

1 (1.5 %)

2 (3.4 %)

 

 nodular

2 (3.1 %)

6 (10.2 %)

 

 punctate

2 (3.1 %)

0 (0 %)

 

 mixed*

1 (1.5 %)

3 (5.1 %)

 

Degree of enhancement

 no/scant

4 (6.2 %)

0 (0 %)

< 0.001

 decreased

29 (44.5 %)

18 (30.5 %)

 

 iso-

25 (38.5 %)

15 (25.4 %)

 

 increased

7 (10.8 %)

26 (44.1 %)

 

Pattern of enhancement

 homogeneous

13 (20 %)

4 (6.8 %)

0.06

 inhomogeneous

52 (80 %)

55 (93.2 %)

 

CT halo sign

 absent

54 (83.1 %)

17 (28.8 %)

< 0.001

 present

11 (16.9 %)

42 (71.2 %)

  1. Note. — *, ‘Mixed’ means punctate plus nodular calcifications. Data are number of items, with percentage in parentheses. CT computed tomography