Endoscopy 2020; 52(S 01): S326
DOI: 10.1055/s-0040-1705053
ESGE Days 2020 ePoster presentations
Thursday, April 23, 2020 09:00 – 17:00 Endoscopic technology ePoster area
© Georg Thieme Verlag KG Stuttgart · New York

USE OF ARTIFICIAL INTELLIGENCE TO IMPROVE POLYP DETECTION RATE

LG Pravia
1   Hospital Clínico San Carlos, Gastroenterology, Madrid, Spain
,
MV Romero
2   Hospital Clinico San Carlos, Madrid, Spain
,
GM Villarejo
2   Hospital Clinico San Carlos, Madrid, Spain
,
FG Rodríguez
2   Hospital Clinico San Carlos, Madrid, Spain
,
JLM Hernández
2   Hospital Clinico San Carlos, Madrid, Spain
,
FS Ceballos
2   Hospital Clinico San Carlos, Madrid, Spain
,
NL Palacios
2   Hospital Clinico San Carlos, Madrid, Spain
,
SF Díez
2   Hospital Clinico San Carlos, Madrid, Spain
,
ER Díaz-Rubio
2   Hospital Clinico San Carlos, Madrid, Spain
,
JM Esteban López-Jamar
2   Hospital Clinico San Carlos, Madrid, Spain
› Author Affiliations
Further Information

Publication History

Publication Date:
23 April 2020 (online)

Also available at
 

Aims Investigate the effect of Artificial Intelligence (AI) on polyp detection rate (PDR).

Methods Seven patients underwent two consecutive colonoscopic examinations by two different endoscopists blind to each other, first one using high definition (HD) and the second one using AI.

Results PDR was 51.14% using HD and 71.42% when adding AI. The mean number of polyps per patient (MPP) was 1.57 with HD (mean size 5.42 mm) and 3.71 when adding AI (mean size 2.8 mm).

Conclusions AI represents a new endoscopic tool that allows to improve PDR as well as to increase MPP, especially of diminutive polyps (< 5 mm).