Artificial Intelligence Improves Novices' Bronchoscopy Performance: A Randomized Controlled Trial in a Simulated Setting.

artificial intelligence assessment feedback flexible bronchoscopy simulation

Journal

Chest
ISSN: 1931-3543
Titre abrégé: Chest
Pays: United States
ID NLM: 0231335

Informations de publication

Date de publication:
23 Aug 2023
Historique:
received: 23 06 2023
revised: 07 08 2023
accepted: 07 08 2023
pubmed: 25 8 2023
medline: 25 8 2023
entrez: 24 8 2023
Statut: aheadofprint

Résumé

Navigating through the bronchial tree and visualizing all bronchial segments is the initial step toward learning flexible bronchoscopy. A novel bronchial segment identification system based on artificial intelligence (AI) has been developed to help guide trainees toward more effective training. Does feedback from an AI-based automatic bronchial segment identification system improve novice bronchoscopists' end-of-training performance? The study was conducted as a randomized controlled trial in a standardized simulated setting. Novices without former bronchoscopy experience practiced on a mannequin. The feedback group (FG, n = 10) received feedback from the AI, and the control group (CG, n = 10) trained according to written instructions. Each participant decided when to end training and proceed to performing a full bronchoscopy without any aids. The FG performed significantly better on all three outcome measures (median difference, P-value): diagnostic completeness (3.5 segments, P < .001), structured progress (13.5 correct progressions, P < .001), and procedure time (-214 seconds, P = .002). Training guided by this novel AI makes novices perform more complete, more systematic, and faster bronchoscopies. Future studies should examine its use in a clinical setting and its effects on more advanced learners.

Sections du résumé

BACKGROUND BACKGROUND
Navigating through the bronchial tree and visualizing all bronchial segments is the initial step toward learning flexible bronchoscopy. A novel bronchial segment identification system based on artificial intelligence (AI) has been developed to help guide trainees toward more effective training.
RESEARCH QUESTION OBJECTIVE
Does feedback from an AI-based automatic bronchial segment identification system improve novice bronchoscopists' end-of-training performance?
STUDY DESIGN AND METHODS METHODS
The study was conducted as a randomized controlled trial in a standardized simulated setting. Novices without former bronchoscopy experience practiced on a mannequin. The feedback group (FG, n = 10) received feedback from the AI, and the control group (CG, n = 10) trained according to written instructions. Each participant decided when to end training and proceed to performing a full bronchoscopy without any aids.
RESULTS RESULTS
The FG performed significantly better on all three outcome measures (median difference, P-value): diagnostic completeness (3.5 segments, P < .001), structured progress (13.5 correct progressions, P < .001), and procedure time (-214 seconds, P = .002).
INTERPRETATION CONCLUSIONS
Training guided by this novel AI makes novices perform more complete, more systematic, and faster bronchoscopies. Future studies should examine its use in a clinical setting and its effects on more advanced learners.

Identifiants

pubmed: 37619664
pii: S0012-3692(23)05276-5
doi: 10.1016/j.chest.2023.08.015
pii:
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Informations de copyright

Copyright © 2023 The Author(s). Published by Elsevier Inc. All rights reserved.

Déclaration de conflit d'intérêts

Financial/Nonfinancial Disclosures The authors have reported to CHEST the following: K. M. C. received partial funding from Ambu regarding The CoRS-feedback study in colonoscopy: NCT04862793. L. K. has annotated clinical bronchoscopy videos for Ambu’s development of the AI system. None declared (S. X., A. O. N., and P. F. C.).

Auteurs

Kristoffer Mazanti Cold (KM)

Copenhagen Academy for Medical Education and Simulation (CAMES), Rigshospitalet, University of Copenhagen and the Capital Region of Denmark. Electronic address: kristoffer.mazanti.cold.01@regionh.dk.

Sujun Xie (S)

Copenhagen Academy for Medical Education and Simulation (CAMES), Rigshospitalet, University of Copenhagen and the Capital Region of Denmark; Guangdong Academy for Medical Simulation (GAMS), Guangzhou, China.

Anne Orholm Nielsen (AO)

Copenhagen Academy for Medical Education and Simulation (CAMES), Rigshospitalet, University of Copenhagen and the Capital Region of Denmark; Herlev University Hospital, Department of Pulmonary Diseases, Herlev, Denmark.

Paul Frost Clementsen (PF)

Copenhagen Academy for Medical Education and Simulation (CAMES), Rigshospitalet, University of Copenhagen and the Capital Region of Denmark.

Lars Konge (L)

Copenhagen Academy for Medical Education and Simulation (CAMES), Rigshospitalet, University of Copenhagen and the Capital Region of Denmark.

Classifications MeSH