ColoRectalCADx: Expeditious Recognition of Colorectal Cancer with Integrated Convolutional Neural Networks and Visual Explanations Using Mixed Dataset Evidence

Table 1

Class labels of the mixed dataset.

Classes	Original mixed dataset	Images	Modified classes	Modified mixed dataset	Images
0	Barretts	42	0	bbps-0-1	653
1	Barretts-short-segment	53	1	bbps-2-3	1148
2	bbps-0-1	653	2	Cecum	2009
3	bbps-2-3	1148	3	Dyed-lifted-polyps	2003
4	Cecum	2009	4	Dyed-resection-margins	1990
5	Dyed-lifted-polyps	2003	5	Esophagitis-a	1404
6	Dyed-resection-margins	1990	6	Non_polyps	818
7	Esophagitis-a	1404	7	Polyps	2150
8	Esophagitis-b-d	260	8	Pylorus	2000
9	Hemorrhoids	10	9	Retroflex-rectum	391
10	Ileum	9	10	Retroflex-stomach	765
11	Impacted-stool	132	11	Ulcerative-colitis-grade-0-1	1035
12	Non_polyps	818	12	Ulcerative-colitis-grade-2	443
13	Polyps	2150	13	-line	1933
14	Pylorus	2000
15	Retroflex-rectum	391
16	Retroflex-stomach	765
17	Ulcerative-colitis-grade-0-1	1035
18	Ulcerative-colitis-grade-1	201
19	Ulcerative-colitis-grade-1-2	11
20	Ulcerative-colitis-grade-2	443
21	Ulcerative-colitis-grade-2-3	28
22	Ulcerative-colitis-grade-3	133
23	-line	1933
		19621			16942