CardioX: Cardiac imaging reports for Precision Cardiology within LTHT
Research type
Research Study
Full title
CardioX: integrated and quantitative cardiac imaging reports for Precision Cardiology within the LTHT
IRAS ID
272645
Contact name
Alejandro Frangi Caregnato
Contact email
Sponsor organisation
University of Leeds
Duration of Study in the UK
0 years, 11 months, 31 days
Research summary
This project aims to improve clinical decision processes, reduce medical costs and act as a test case for the adoption of AI-enabled solutions at the Leeds Teaching Hospitals NHS Trust and more broadly across the NHS.
Cardiovascular disease is number one cause of death within the UK. Early quantitative assessment of cardiac function is crucial for proper diagnosis and treatment of cardiovascular disease. Leeds has one of the largest cardiac MRI services in the UK conducting approximately 3,000 cardiovascular MRI scans within the region, at an annual cost of £1.38M per year. Manually extracting quantitative information from these MRI studies is a time consuming, laborious process and may lead to inter-observer variability. A senior cardiologist requires approximately 45-60 min to interpret and report each study. This project will automate the analysis of MRI scans and the generation of personalised reports providing in a fraction of that time a comprehensive set of quantitative measures characterising cardiac functions.
The automated contouring and report generation is not designed to replace manual interventions. Clinicians will be able to review and amend all aspects of the contours and reports or compare them with their entirely manually generated MRI reviews. The results of their adjustments will be used to further inform and enhance the reliability of the tool and be used to validated it’s accuracy. The automated analysis will be further enhanced through the inclusion of additional relevant contextual data that is held within the patient’s electronic health records. This will enable the identification of patterns of association between cardiac function, structural cardiac defects and comorbidities, demographics, lifestyle, or other clinical conditions.
The study will be carried out securely within the NHS by LTHT personnel. De-identified data will be extracted from the hospital’s electronic health record called PPM. This data will include the raw results of MRI data, their associated clinical reports and contours, and other clinically relevant structured data. All data will be extracted and anonymised by a team separate to the research team prior to the release of the research dataset. The study will include all adult patient’s that have undergone a cardiac MRI within LTHT except those who have registered to opt out of data based research. Anonymised data will be extracted from PPM the electronic patient record developed by LTHT and in use across Yorkshire.REC name
East of England - Cambridge Central Research Ethics Committee
REC reference
19/EE/0373
Date of REC Opinion
12 Dec 2019
REC opinion
Favourable Opinion