Table 1 Dataset information
From: Automated detection of pulmonary embolism from CT-angiograms using deep learning
| Â | Training set | Test set |
|---|---|---|
CTPAs (stacks) | 608 | 204 |
 Positive | 303 (50%) | 97 (48%) |
 Negative | 305 (50%) | 107 (52%) |
CT slices | 52,752 | 17,778 |
 Positive | 7170 (14%) | 2801 (16%) |
 Negative | 45,582 (86%) | 14,977 (84%) |
Distinct patients | 569 | 201 |
 Male | 250 (44%) | 88 (44%) |
 Female | 319 (56%) | 113 (56%) |
Mean age | 64 | 64 |
CT manufacturer | ||
 Toshiba | 569 (94%) | 195 (96%) |
 GE Medical Systems | 21 (3%) | 5 (2%) |
 Siemens | 18 (3%) | 4 (2%) |