Automated Diagnosis of Pulmonary Embolism on CTPA

  • Research type

    Research Study

  • Full title

    Developing a deep learning model for the automated detection of pulmonary embolism on computed tomography pulmonary angiogram.

  • IRAS ID

    341910

  • Contact name

    Stephanie Lua

  • Contact email

    stephanie.lua@ggc.scot.nhs.uk

  • Sponsor organisation

    NHS Golden Jubilee National Hospital

  • Duration of Study in the UK

    1 years, 0 months, 1 days

  • Research summary

    This study aims to develop a deep learning model using artificial intelligence (AI) that can independently detect pulmonary thromboembolism (clots in the lungs) on CT scans.
    This AI model can be run on a standard CT scan obtained in patients who are undergoing normal treatment for their condition ( standard of care), and does not require any additional radiation or scanning requirements.
    This will be performed in conjunction with the medical research arm of Canon Medical (a company which makes CT scanners).

    The Canon engineers will develop a computer model trained on publicly available CT scans, and we will validate this model on retrospective scans of patients with known clots in the lungs, under the care of the Scottish Pulmonary Vascular Unit. This data will be anonymized prior to transferring to Canon Medical.

    We aim to compare the results of the scans between an official radiology report, and the AI model, to calculate the sensitivity and specificity of this model.

  • REC name

    London - Hampstead Research Ethics Committee

  • REC reference

    24/PR/1481

  • Date of REC Opinion

    16 Jan 2025

  • REC opinion

    Further Information Favourable Opinion