NEW-AGE Study V 1.0

  • Research type

    Research Study

  • Full title

    Collection of Endoscopic Images and Videos for Validation of Artificial Intelligence Algorithms for Computer Aided Detection and Characterization of Early Gastrointestinal Neoplasia

  • IRAS ID

    289121

  • Contact name

    Pradeep Bhandari

  • Contact email

    pradeep.bhandari@porthosp.nhs.uk

  • Sponsor organisation

    Portsmouth Hospitals NHS University Trust

  • Duration of Study in the UK

    2 years, 0 months, 0 days

  • Research summary

    Gastrointestinal (GI) neoplasia is a major health problem. Endoscopy remains the gold standard test for early detection and therapy. However, despite all recent advances in imaging techniques, there is still a significant neoplasia miss rate contributing to missed pathologies and interval cancer. Artificial Intelligence (AI) is proposed to play a role in computer aided detection and characterization of early GI neoplasia and early research in this area is encouraging.

    In this study, we aim to collect anonymized endoscopic data (images and videos) and use them to validate an existing AI algorithm and test its performance. Anonymous endoscopic images and videos are routinely recorded and collected during endoscopy procedures, and all patients will have been consented for that as part of the standard endoscopy consent process.

    In the initial phase of the study, we will use previously collected and fully anonymous endoscopic images and videos from our departmental teaching and training library and use them for validation and optimization of the computer algorithm. In the second phase, we will prospectively collect anonymous endoscopic videos recorded during endoscopy procedures and use them to test the performance of the computer algorithm. Therefore, this study will have no impact on patient’s routine care or treatment.

    Validation and testing of the AI algorithm using anonymous endoscopic images and videos will take place at the Sponsor site, Queen Alexandra Hospital, with no interference with routine patient’s care.

  • REC name

    London - Queen Square Research Ethics Committee

  • REC reference

    20/PR/0985

  • Date of REC Opinion

    28 Jan 2021

  • REC opinion

    Further Information Favourable Opinion