Table 1 Statistics of our constructed datasets from real-world EMRs in two Chinese hospitals
From: Optimising the paradigms of human AI collaborative clinical coding
HPH-50 | APH-50 | HPH-100 | APH-100 | HPH-full | APH-full | |
|---|---|---|---|---|---|---|
Number of | ||||||
# Doc. | 10,223 | 9514 | 10,682 | 14,104 | 10,682 | 14,104 |
Avg # words per Doc. | 728 | 1472 | 725 | 1051 | 725 | 1051 |
Avg # codes per Doc. | 4.27 | 4.03 | 4.89 | 4.78 | 5.97 | 5.58 |
Total # codes | 50 | 50 | 100 | 100 | 671 | 579 |
Completeness ratio of | ||||||
# Specialist condition | 65.27% | 60.15% | 62.27% | 60.51% | 62.27% | 60.51% |
# Auxiliary examination | 47.76% | 83.67% | 47.13% | 63.04% | 47.13% | 63.04% |
