Multivariable Lung Nodule Malignancy Prediction
Research type
Research Study
Full title
Multivariable prediction of cancer in patient records with lung nodules found on chest CT performed in routine clinical practice: a retrospective population-wide cohort study
IRAS ID
316195
Contact name
Huan Wang
Contact email
Sponsor organisation
University of Dundee
Duration of Study in the UK
1 years, 4 months, 31 days
Research summary
This project aims to evaluate the impact of an artificial intelligence (AI) tool for the detection of lung nodules, which may develop into a cancerous growth. We know from previous publications that the risk of these nodules is not uniform, and some nodule characteristics show an increased risk for cancer development.
To gain the necessary information, we plan to investigate retrospective data provided by the Scottish Medical Imaging (SMI) service, run by Public Health Scotland, which comprises a large number of routinely collected computed tomography (CT) lung scans and link the imaging findings with essential clinical data, including the diagnosis of lung cancer, risk factors in patients, and general information like sex and age. This will result in a risk score, which will then be linked to the CT findings. It is anticipated that the resulting artificial intelligence (AI) tool, comprising prediction models, will improve the early detection of lung cancer once put into production as a certified device. CE certification will be undertaken separately and is out of scope for this study.
REC name
North of Scotland Research Ethics Committee 1
REC reference
23/NS/0028
Date of REC Opinion
2 Apr 2023
REC opinion
Favourable Opinion