Skip to main content

Table 1 Statistics of discharge summaries in the MIMIC-III training data

From: Improving medical term embeddings using UMLS Metathesaurus

# training notes

47,423

# of unique medical terms in training data

46,861

Average # of medical terms in a discharge summary

671

# of unique medical concepts in training data

29,740

Average # of medical concepts per discharge summary

364

Average # of definition words per medical concept

16

# of unique diagnosis codes in training data

6717

Average # of diagnosis codes per discharge summary

11