Table 21 Analysis of ADaDR-22 + Vision transformer in depth (Modified dataset).

From: Exploring vision transformers and XGBoost as deep learning ensembles for transforming carcinoma recognition

Class

Accuracy (%)

F1 Score (%)

Recall (%)

Precision (%)

Support

BBPS-0–1

93.1

93.1

93

93.2

500

BBPS-2–3

94

94

94.1

93.9

450

Cecum

93.8

93.8

93.7

93.9

700

Dyed-Lifted-Polyps

94.2

94.2

94.3

94.1

650

Dyed-Resection-Margins

93.5

93.5

93.6

93.4

600

Esophagitis-A

93.7

93.7

93.8

93.6

550

Non-Polyps

94.1

94.1

94

94.2

520

Polyps

93.9

93.9

94

93.8

800

Pylorus

94

94

94.1

93.9

480

Retroflex-Rectum

93.3

93.3

93.2

93.4

300

Retroflex-Stomach

94.1

94.1

94

94.2

490

Ulcerative-Colitis-Grade-0–1

94.3

94.3

94.4

94.2

430

Ulcerative-Colitis-Grade-2

93.4

93.4

93.3

93.5

320

Z-Line

94.2

94.2

94.3

94.1

600