Table 1:

Training and test set characteristics

Set	CASE	NON-CASE	Total	Inter-annotator agreement (κ)
Training	132	348	480	Single annotated
Test	38	82	120	0.828