UEG's content platform for gastroenterology education bringing recordings, courses, guidelines, and more together in one place.

Re-live UEG Week 2026 on Gutflix Gut Guide online What's new on Gutflix

Trending on Gutflix

Rate for better recommendations!

Your feedback helps us show you more of what you like and less of what you dislike!

More of this!

We will show you more similar content.

Not for me!

We won't suggest this again.

Bookmark!

Save this content for later!

New to myUEG? Create an account! Log In with myUEG account

UEG Poster

Standard Poster

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

CLINICAL VALIDATION OF POLYDEEP, A COMPUTER-AID DIAGNOSIS SYSTEM: RANDOMISED MULTICENTRE CLINICAL TRIAL WITH TANDEM COLONOSCOPY DESIGN

Pedro Davila-Piñón 1,2, Cristina Regueiro Expósito 1,2, Astrid Irene Díez Martín 1,2, Jorge Hernández-Camoiras 1,2, Sara Zarraquiños Martínez 3, Maria Luisa De Castro Parga 4, Jesús Herrero 3, Nereida Fernández 4, Pablo Vega 3, David Remedios 3, Alfonso Martínez 4, Manuel Puga 3, Sara Alonso 4, Noel Pin Vieito 3, Natalia García-Morales 4, Laura Rivas 3, Alejandro Ledo 5, RAMIRO MACENLLE 3, Lucia Cid 4, Antonio Rodríguez 4, Santiago Soto 3, Franco Baiocchi 3, Indhira Miosotis 5, Eloy Sánchez 3, Rubén Domínguez-Carbajales 3, Alba Nogueira-Rodríguez 6,7, Hugo López-Fernández 6,7, Florentino Fdez-Riverola 6,7, Daniel Glez-Peña 6,7, Miguel Reboiro-Jato 6,7, Joaquín Cubiella 1,3

Affiliations

1 Research Group in Gastrointestinal Oncology, Ourense, Spain

2 Galicia Sur Biomedical Foundation, Ourense, Spain

3 Complexo Hospitalario Universitario de Ourense, Ourense, Spain

4 Hospital Álvaro Cunqueiro de Vigo - CHUVI, Vigo, Spain

5 Hospital Montecelo - CHUP, Pontevedra, Spain

6 CINBIO, Universidade de Vigo, ESEI-Escuela Superior de Ingeniería Informática, Ourense, Spain

7 SING Research Group., Galicia Sur Health Research Institute (IIS Galicia Sur), Ourense, Spain

Summary

AI Generated

In a multicentric randomised tandem colonoscopy trial, the AI-based Computer-Aided Detection system PolyDeep did not reduce adenoma miss rate when used by experienced endoscopists in a colorectal cancer screening population.

The PolyDeep Advance 2 trial randomised 244 patients (after exclusions) to first colonoscopy with or without PolyDeep assistance in a tandem colonoscopy design, with the primary endpoint of adenoma miss rate.
Adenoma miss rate was 21.6 percent with PolyDeep assistance versus 17.0 percent without assistance (p=0.2), showing no statistically significant difference.
No statistically significant differences were observed for any secondary endpoints, including serrated lesion miss rate (23.4 percent vs 24.0 percent), polyp miss rate (22.0 percent vs 19.0 percent), or advanced lesion miss rate (10.9 percent vs 10.4 percent).
The study population included patients with positive faecal immunochemical test or requiring surveillance after advanced adenoma resection, with 727 colonic lesions detected (66.4 percent adenomas, 23.8 percent serrated lesions).
This trial is relevant to gastroenterologists and endoscopists evaluating AI-assisted colonoscopy technologies in expert practice settings.

Download PDF

Was this helpful?

Thanks for your feedback.

This summary was generated by an AI large language model based on the content transcript. It is for informational purposes only and should not be considered a substitute for clinical judgment. Always rely on your professional expertise and the full clinical context when making clinical decisions.

Abstract

Introduction

Computer-Aided polyp Detection (CADe) systems based on artificial intelligence (AI) can increase the detection rate of colonic lesions. PolyDeep is a CADe that has demonstrated its ability to detect lesions.

Aims & Methods

The aim of the study is to evaluate its performance in the context of expert endoscopists in a population-based colorectal cancer screening programme. PolyDeep Advance 2 (NCT05512793) is a multicentric randomised clinical trial with a tandem colonoscopy design. The primary endpoint of this study is to determine and compare whether PolyDeep-assisted colonoscopy reduces the Adenoma Miss Rate (AMR). The secondary outcomes are to determine and compare the serrated lesion miss rate, the polyp miss rate including adenomas and serrated lesions in this category, the advanced lesion miss rate (adenomas and serrated lesions) and the small or diminutive lesion miss rate. We included patients referred for colonoscopy with a positive faecal inmunochemical test result, or who required surveillance after resection of advanced adenomas. All baseline, patient, procedure and lesion characteristics were compared using chi-squared and t-Student tests.

Results

Between May 2023 and November 2023 we included 260 patients who were randomised to first colonoscopy with PolyDeep (130) or first colonoscopy without PolyDeep (130). We excluded 16 patients due to lack of double cecal intubation (15) or poor bowel preparation (1). Baseline characteristics of the 244 patients (124 first PolyDeep-assisted colonoscopy, 120 first non-assisted colonoscopy) were equivalent (male: 62.3%, years of age: 62.3 ± 6.5, FIT positive: 66.0%, mean Boston Bowel Preparation Scale: 7.5 ± 1.3, first withdrawal time: 13:47 ± 08:34 min, second withdrawal time: 07:52 ± 03:39 min, detection rate of colonic lesions: 77.0%, mean number of colorectal lesions per colonoscopy: 3.0 ± 4.3). We detected 727 colonic lesions (66.4% adenomas, 23.8% serrated lesions, 6.1% other histologies, 3.7% non-retrieved lesions). We found no statistically significant differences in the miss rates between groups (Table).

	PolyDeep-assisted colonoscopy	Non-assisted colonoscopy	p
Adenoma Miss Rate	21.6%	17.0%	0.2
Serrated Lesion Miss Rate	23.4%	24.0%	0.9
Polyp Miss Rate	22.0%	19.0%	0.2
Advanced Lesion Miss Rate	10.9%	10.4%	0.9
Small lesion Miss Rate	23.4%	21.0%	0.5
Diminutive lesion Miss Rate	25.5%	21.6%	0.3

Conclusion

In colonoscopies performed by experienced endoscopists, PolyDeep does not reduce the AMR.

Conference

UEG Week Vienna 2024

Topics

Colorectal

Submission format

Abstract

Session

Marching forward in lower GI endoscopic treatment (Posters)

Citation

United European Gastroenterology Journal 2024; 12 (Supplement 8)

Published

2024

More Like This:

POLYDEEP ADVANCE 3: MULTICENTRE RANDOMIZED CONTROLLED TRIAL COMPARING THE POLYDEEP-ASSISTED COLONOSCOPY WITH CONVENTIONAL COLONOSCOPY

María José Iglesias Varela

CLINICAL VALIDATION OF POLYDEEP A COMPUTER-AIDED DIAGNOSIS SYSTEM (CADE/X)

Hugo López-Fernández

OPTICAL DIAGNOSIS OF COLORECTAL LESIONS: NESTED ANALYSIS OF THREE CLINICAL TRIALS COMPARING POLYDEEP CADX WITH EXPERT ENDOSCOPISTS

Joaquín Cubiella

COMPARISON OF THE DIAGNOSTIC ACCURACY OF POLYDEEP, AN ARTIFICIAL INTELLIGENCE COMPUTER-AIDED DIAGNOSIS SYSTEM (CADE/X) WITH ENDOSCOPISTS FOR THE CHARACTERIZATION OF COLORECTAL POLYPS

Astrid Irene Díez Martín

THE CLINICAL VALIDATION OF A COMPUTER-AIDED POLYP DETECTION MODEL INTEGRATED AS A PLUG-AND-PLAY ENDOSCOPY DEVICE (ALTER-EGO TRIAL)

Raf Bisschops

WHAT FACTORS ARE ASSOCIATED WITH THE OPTICAL DIAGNOSIS OF POLYDEEP, A COMPUTER-AIDED DIAGNOSIS SYSTEM (CADE/X)? IN VITRO ANALYSIS

Astrid Irene Díez Martín

UEG Presentation

Share via Email Share on Facebook Share on X Share on LinkedIn Share on Bluesky

POLYDEEP ADVANCE 3: MULTICENTRE RANDOMIZED CONTROLLED TRIAL COMPARING THE POLYDEEP-ASSISTED COLONOSCOPY WITH CONVENTIONAL COLONOSCOPY

This content is part of Gutflix. Log in with your myUEG account, or create one free, to watch it.

Not sure what you can access? Learn more about account types.

POLYDEEP ADVANCE 3: MULTICENTRE RANDOMIZED CONTROLLED TRIAL COMPARING THE POLYDEEP-ASSISTED COLONOSCOPY WITH CONVENTIONAL COLONOSCOPY

Pedro Davila-Piñón 1,2, Astrid Irene Díez Martín 1,2, Cristina Regueiro Expósito 1,2, Alba Nogueira-Rodríguez 3,4, Florentino Fdez-Riverola 3,4, Miguel Reboiro-Jato 3,4, Daniel Glez-Peña 3,4, Rubén Domínguez-Carbajales 5, Luisa De Castro Parga 6, Jesús Daniel Fernández de Castro 7, Pablo Vega 7, Miguel Telmo Galovart Araguas 6, Santiago Soto 7, Sara Alonso Lorenzo 6, Alfonso Martinez Turnes 6, Noel Pin Vieito 7, Sara Zarraquiños Martínez 7, David Remedios 7, Cristina Alejandra Sánchez Gómez 7, Raquel Souto Rodríguez 8, Franco Baiocchi 7, María José Iglesias Varela 7, Coral Tejido Sandoval 7, Alejandro Ledo 8, Laura Rivas Moral 7, Natalia Garcia Morales 6, Antonio Rodríguez 6, Manuel Puga 7, María Belén Castiñeira-Domínguez 8, Nereida Fernández Fernández 6, Arantxa Germade-Martínez 6, Enrique González 8, Indhira Miosotis 8, Romina Fernández Poceiro 6, Eloy Sanchez Hernandez 7, Hugo López-Fernández 3,4, Joaquín Cubiella 1,7

Affiliations

1 Research Group in Gastrointestinal Oncology, Ourense, Spain

2 Galicia-Sur Public Biomedical Foundation, Ourense, Spain

3 Universidade de Vigo, Ourense, Spain

4 SING Research Group, Ourense, Spain

5 Complexo Hospitalario Universitario de Ourense, Ourense, Spain

6 Hospital Álvaro Cunqueiro de Vigo - CHUVI, Vigo, Spain

7 Hospital Universitario de Ourense, Ourense, Spain

8 Hospital Montecelo - CHUP, Pontevedra, Spain