AID-MR: Artificial Intelligence Delivered Cardiac Magnetic Resonance - Prospective Validation
Study Details
Study Description
Brief Summary
Cardiac MRI (CMR) scanning allows doctors to create detailed images of the heart. However, the need for experienced cardiac radiographers to perform each scan can make CMR's delivery difficult, and some patients in the UK wait more than half a year for a scan. These radiographers must take pictures of different part of the heart, termed "views", each of which must be precisely positioned.
The investigators believe they can revolutionise CMR, by using artificial intelligence to automatically position the views so radiographers can focus on more difficult tasks.
The investigators have used a retrospective database of pseudonymised (anonymised and linked) CMR scans at our hospital to create these artificial intelligence (AI) algorithms, and they have validated them retrospectively on previous studies. The investigators now wish to test the algorithms prospectively.
In this study, the investigators will recruit patients undergoing clinical CMR scans. In addition to the routine images acquired by expert radiographers, the investigators will require a duplicate set of images, positioned and planned by the AI algorithms.
The investigators will then compare, within each patient, the AI-planned and expert-radiographer-planned scanning in terms of both speed and image quality.
| Condition or Disease | Intervention/Treatment | Phase |
|---|---|---|
|
N/A |
Study Design
Arms and Interventions
| Arm | Intervention/Treatment |
|---|---|
| Experimental: AI-planned images acquired
|
Diagnostic Test: AI-assisted cardiac magnetic resonance imaging
An AI algorithm will be used to automatically position (plan) the scan planes used in a cardiac MRI scan. The resultant images will be compared with standard radiographer-positioned images.
|
| Active Comparator: Radiographer-planned images acquired
|
Diagnostic Test: AI-assisted cardiac magnetic resonance imaging
An AI algorithm will be used to automatically position (plan) the scan planes used in a cardiac MRI scan. The resultant images will be compared with standard radiographer-positioned images.
|
Outcome Measures
Primary Outcome Measures
- Time taken to acquire images [During the MRI scan]
The time taken in seconds from the beginning of the planning process, until the last planned images has been fully acquired.
- Image quality [During the MRI scan]
Quality of acquired images (AI-planning versus radiographer planning) assessed by level 3 cardiac MRI accredited doctors, by asking them to choose whether (a) the AI-acquired image is of higher diagnostic quality, (b) the radiographer-acquired image is of higher diagnostic quality, or (c) the AI- and radiographer-acquired images are of identical diagnostic quality.
Eligibility Criteria
Criteria
Inclusion Criteria:
- Adult (aged at least 18 years)
Exclusion Criteria:
-
Children (patients below age 18).
-
Pregnant patients.
Contacts and Locations
Locations
No locations specified.Sponsors and Collaborators
- Imperial College London
- British Heart Foundation
- Medical Research Council
- Rosetrees Trust
Investigators
- Principal Investigator: James P Howard, MB BChir PhD, Imperial College London
Study Documents (Full-Text)
None provided.More Information
Publications
None provided.- 23HH8238