Table 2 Patients’ characteristics.
From: Deep learning-based diagnosis from endobronchial ultrasonography images of pulmonary lesions
|  | Training dataset (N = 171) | Test dataset (N = 42) | P-value |
|---|---|---|---|
Age, year, mean | 72.7 ± 10.8 | 74.8 ± 12.6 | 0.072 |
Sex, female (%)/male (%) | 57 (33)/114 (77) | 12 (29)/30 (71) | 0.5873 |
Lesion size, cm, mean | 3.5 ± 2.1 | 3.1 ± 1.9 | 0.1856 |
Endobronchial ultrasound visualization, Adjacent to (%)/Within (%) | 35 (20)/136 (80) | 12 (29)/30 (71) | 0.2992 |
Stage (%) | Â | Â | 0.8394 |
    IA-B | 51 (40) | 12 (40) |  |
    IIA-B | 9 (7) | 1 (3) |  |
    IIIA-C | 20 (16) | 4 (13) |  |
    IVA-B | 47 (37) | 13 (44) |  |
