Table 2 Patient and CXR acquisition characteristics for each dataset. Data is shown in absolute number and percentage in parenthesis. Age is shown as median [minimum; maximum]. * indicates that the information could not be obtained for all subjects and NA indicates that this data is not available.
From: Assessing clinical applicability of COVID-19 detection in chest radiography with deep learning
Mixed | BIMCV | COVIDGR | CHVNGE | |
|---|---|---|---|---|
Gender | ||||
Male | 165,123 (58.0) | 8615 (51.8) | 0 (0.0) | 3340 (52.5) |
Female | 118,863 (41.7) | 8015 (48.2) | 0 (0.0) | 3021 (47.5) |
Unknown/Other | 807 (0.3) | 1 (0.0) | 852 (100.0) | 0 (0.0) |
Age | 58 [1;106]* | 64 [1;101]* | NA | 66 [2;100] |
View | ||||
AP | 199,606 (70.1) | 4,214 (25.3) | 0 (0.0) | 82 (1.3) |
PA | 84,070 (29.5) | 12,160 (73.1) | 852 (100.0) | 66 (1.0) |
Unknown | 1117 (0.4) | 257 (1.5) | 0 (0.0) | 6213 (97.7) |
CXR equipment | ||||
Agfa CR | 2280 (0.8) | 346 (2.1) | 0 (0.0) | 0 (0.0) |
Agfa DX | 247 (1.5) | 0 (0.0) | 0 (0.0) | |
Canon DX | 0 (0.0) | 45 (0.3) | 0 (0.0) | 0 (0.0) |
Carestream CR | 0 (0.0) | 9 (0.1) | 0 (0.0) | 0 (0.0) |
Carestream DX | 0 (0.0) | 97 (0.6) | 0 (0.0) | 59 (0.9) |
FUJI CR | 0 (0.0) | 155 (0.9) | 0 (0.0) | 2164 (34.0) |
FUJI DX | 0 (0.0) | 4 (0.0) | 0 (0.0) | 436 (6.9) |
GE DX | 0 (0.0) | 155 (0.9) | 0 (0.0) | 0 (0.0) |
GMM DX | 0 (0.0) | 269 (1.6) | 0 (0.0) | 0 (0.0) |
ImagingDynamics CR | 0 (0.0) | 3668 (22.1) | 0 (0.0) | 0 (0.0) |
KONICA CR | 0 (0.0) | 423 (2.5) | 0 (0.0) | 0 (0.0) |
KONICA DX | 0 (0.0) | 89 (0.5) | 0 (0.0) | 0 (0.0) |
Philips CR | 1019 (0.4) | 8131 (48.9) | 0 (0.0) | 0 (0.0) |
Philips DX | 2711 (16.3) | 0 (0.0) | 0 (0.0) | |
Samsung DX | 0 (0.0) | 0 (0.0) | 0 (0.0) | 3702 (58.2) |
SIEMENS CR | 1575 (0.6) | 261 (1.6) | 0 (0.0) | 0 (0.0) |
SIEMENS DX | 0 (0.0) | 0 (0.0) | 0 (0.0) | |
Unknown/Other | 279,919 (98.3) | 21 (0.1) | 852 (100.0) | 0 (0.0) |
