Skip to main content

Table 4 Subgroup analysis of the LLFs in danger zone nodule

From: The diagnostic performance and clinical value of deep learning-based nodule detection system concerning influence of location of pulmonary nodule

 

DLD system

p-value

Thoracic radiologists

p-value

Non-thoracic radiologists

p-value

Radiology residents

p-value

Location

Apical lung zone

87.9 (29/33)

0.002*

84.9 (56/66)

0.319

69.7 (46/66)

0.032*

65.9 (87/132)

0.120

Paramediastinal

52.5 (31/59)

78.8 (93/118)

53.4 (63/118)

57.6 (136/236)

Retrodiaphragmatic

58.8 (10/17)

85.3 (29/34)

41.2 (14/34)

35.3 (24/68)

Visibility score

 1

16.7 (1/6)

0.169

33.3 (4/12)

0.129

16.7 (2/12)

0.271

20.8 (5/24)

0.066

 2

50.0 (12/24)

58.3 (28/48)

33.3 (16/48)

41.7 (40/96)

 3

64.4 (38/59)

90.7 (107/118)

60.2 (71/118)

58.5 (138/236)

 4

95.0 (19/20)

97.5 (39/40)

85.0 (34/40)

80.0 (64/80)

Size

 ≤ 2 cm

48.5 (33/68)

< 0.001*

74.3 (101/136)

< 0.001*

47.1 (64/136)

< 0.001*

47.8 (130/272)

< 0.001*

 > 2 cm

90.2 (37/41)

93.9 (77/82)

71.9 (59/82)

71.3 (117/164)

  1. DLD Deep learning-based nodule detection, LLF Lesion localization fraction
  2. *p < 0.05 was regarded as statistically significant. Denominator for the LLF calculation corresponds to the total number of true lesions multiplied by the number of readers in the respective group