Table 1 PHI distributions in the 2014 i2b2/UTHealth de-identification corpus and UF Health clinical notes
From: A study of deep learning methods for de-identification of clinical notes in cross-institute settings
PHI Category | Number of Annotations | ||||
|---|---|---|---|---|---|
2014 i2b2/UTHealth | UF Heath | ||||
Training | Validation | Training | Validation | Test | |
DATE | 9067 | 3104 | 2056 | 774 | 1872 |
NAME | 5472 | 1868 | 856 | 356 | 771 |
AGE | 1507 | 490 | 158 | 86 | 164 |
ID | 1142 | 364 | 156 | 41 | 137 |
PHONE | 406 | 128 | 50 | 28 | 47 |
WEB | 6 | 1 | 0 | 0 | 4 |
INSTITUTE | 1926 | 592 | 128 | 72 | 119 |
STREET | 280 | 72 | 25 | 6 | 21 |
CITY | 502 | 152 | 43 | 26 | 45 |
ZIP | 276 | 76 | 34 | 11 | 20 |
Total | 20,584 | 6847 | 3506 | 1400 | 3200 |