Background We are developing artificial intelligence based polyp histology prediction (AIPHP) method to automatically classify Narrow Band Imaging (NBI) magnifying colonoscopy images to predict the non-neoplastic or neoplastic histology of polyps.

Aim Our aim was to analyse the accuracy of AIPHP and NICE classification based histology predictions and also to compare the results of the two methods.

Methods We examined colorectal polyps obtained from colonoscopy patients who had polypectomy or endoscopic mucosectomy. Polyps detected by white light colonoscopy were observed then by using NBI at the optical maximum magnificent (60x). The obtained and stored NBI magnifying images were analysed by NICE classification and by AIPHP method parallelly. Pathology examinations were performed blinded to the NICE and AIPHP diagnosis, as well. Our AIPHP software is based on a machine learning method. This program measures five geometrical and colour features on the endoscopic image.

Inclusion Criteria:

• endoscopic diagnosis of colorectal polyp

Exclusion Criteria:

• colonoscopy result without polyps or IBD diagnosis

patients with symptoms of colorectal polypoid laesions, colorectal cancer screening patients