Skip to main content

Table 1 False positive images & false negative images analysis in the image test set

From: A newly developed deep learning-based system for automatic detection and classification of small bowel lesions during double-balloon enteroscopy examination

 

Misjudged reasons

Number of images

False positive images

Normal mucosa

53

Mucus & faeces

18

Light spot & dim & bubbles

24

Dark view

3

False negative imagesa

Diverticulum

33

Erosion & Ulcer

15

Protruding lesion

19

Angioectasia

5

  1. a Most of the false negative images were misjudged by the following factors: dark view, lesions that are similar to surrounding mucosa and few of lesions with small flat hematin base