Skip to main content

Table 1 Abnormal annotation distribution

From: Temporal bone radiology report classification using open source machine learning and natural langue processing libraries

Region Training set Test set
At least one 62.41 % (362) 62.33 % (91)
Inner 26.72 % (155) 26.03 % (38)
Middle 37.59 % (218) 40.41 % (59)
Outer 13.79 % (80) 14.38 % (21)
Mastoid 30.86 % (179) 36.99 % (54)
  1. Column values indicate the percentage of documents (values in parenthesis indicate absolute number of documents) that were labeled as abnormal for the given region. A document as a whole was considered abnormal if it contained an abnormality in at least one region. The training and test sets contain a total of 580 and 146 documents, respectively