BioMed Research International / 2021 / Article / Tab 2

Research Article

An Interpretable Model-Based Prediction of Severity and Crucial Factors in Patients with COVID-19

Table 2

Chest CT features of the patients.

Mild group ()Severe group ()

Stage0.208
 Early stage44 (27.16%)10 (27.78%)
 Progress stage105 (64.81%)26 (72.22%)
 Restoration stage13 (8.02%)0
Location0.002
 Subpleural50 (30.86%)2 (5.56%)
 Scatter54 (33.33%)11 (30.56%)
 Diffuse58 (35.80%)23 (63.89%)
Shape<0.001
 Nodular11 (6.79%)1 (2.78%)
 Patchy132 (81.48%)17 (47.22%)
 Large patchy19 (11.73%)18 (50.00%)
Number of lobes involved<0.001
 122 (13.58%)0
 222 (13.58%)1 (2.78%)
 320 (12.35%)1 (2.78%)
 434 (20.99%)5 (13.89%)
 564 (39.51%)29 (80.56%)
Image manifestations
 Pleural effusion1 (0.62%)4 (11.11%)0.004
 Fibrosis64 (39.51%)15 (41.67%)0.811
 Consolidation85 (52.47%)28 (77.78%)0.006
 Reticular shadow95 (58.64%)34 (94.44%)<0.001
 Crazy paving9 (5.56%)15 (41.67%)<0.001
 Air bronchogram55 (33.95%)26 (72.22%)<0.001
 Pleural thickening62 (38.27%)24 (66.67%)0.002
 Lymphadenovarix10 (6.17%)4 (11.11%)0.493
 GGO162 (100.00%)36 (100.00%)
 Nodules68 (41.98%)19 (52.78%)0.211
Quantitative features
 CT from onset (days)0.002
 Total score<0.001
 UOR<0.001
 MOR<0.001
 IOR<0.001
 UOL<0.001
 IOL<0.001

GGO: ground-glass opacity; UOR: upper lobe of right lung; MOR: middle lobe of right lung; IOR: inferior lobe of right lung; UOL: upper lobe of left lung; IOL: inferior lobe of left lung.

Article of the Year Award: Outstanding research contributions of 2020, as selected by our Chief Editors. Read the winning articles.